Publications
This bibliography is extracted from various primary sources using automatic language understanding tools. A good faith effort has been made to eliminate errors and minimize omissions. Please bring any remaining errors or omissions to the attention of CLSP by writing to [email protected].
- Trevor Meyer, Camden Shultz, N. Dehak, L. Moro-Velázquez, and Pedro P. Irazoqui, “Time Scale Network: A Shallow Neural Network For Time Series Data,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{265128667, title = {Time Scale Network: A Shallow Neural Network For Time Series Data}, author = {{Trevor Meyer} and {Camden Shultz} and {N. Dehak} and {L. Moro-Velázquez} and {Pedro P. Irazoqui}}, year = 2023, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/deacbb4906e1d2e597602a65b434a8132953ad8d}, } - Helin Wang, Venkatesh Ravichandran, Milind Rao, Becky Lammers, Myra Sydnor, Nicholas J Maragakis, A. Butala, Jayne Zhang, Lora Clawson, Victoria Chovaz, and L. Moro-Velázquez, “Improving fairness for spoken language understanding in atypical speech with Text-to-Speech.” 2023.
[BibTeX] [Link]@inproceedings{265281100, title = {Improving fairness for spoken language understanding in atypical speech with Text-to-Speech}, author = {{Helin Wang} and {Venkatesh Ravichandran} and {Milind Rao} and {Becky Lammers} and {Myra Sydnor} and {Nicholas J Maragakis} and {A. Butala} and {Jayne Zhang} and {Lora Clawson} and {Victoria Chovaz} and {L. Moro-Velázquez}}, year = 2023, month = {11}, booktitle = {}, url = {https://www.semanticscholar.org/paper/45e3115df40de802f9c4095f329ea374aac56825}, } - Karan Thakkar, Jiarui Hai, and Mounya Elhilali, “Investigating Self-Supervised Deep Representations for EEG-based Auditory Attention Decoding,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{264935647, title = {Investigating Self-Supervised Deep Representations for EEG-based Auditory Attention Decoding}, author = {{Karan Thakkar} and {Jiarui Hai} and {Mounya Elhilali}}, year = 2023, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/57b903a709dc9ab64b3bd81378f33547eabb01bd}, } - Guanghui Qin and Benjamin Van Durme, “Nugget: Neural Agglomerative Embeddings of Text,” in International Conference on Machine Learning, 2023.
[BibTeX] [Link]@inproceedings{260957214, title = {Nugget: Neural Agglomerative Embeddings of Text}, author = {{Guanghui Qin} and {Benjamin Van Durme}}, year = 2023, month = {10}, booktitle = {International Conference on Machine Learning}, url = {https://www.semanticscholar.org/paper/531b37c44c7e39539f617fb1a4149ef8cce8f4ec}, } - Amir Feder, Yoav Wald, Claudia Shi, S. Saria, and David Blei, “Causal-structure Driven Augmentations for Text OOD Generalization,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{264305897, title = {Causal-structure Driven Augmentations for Text OOD Generalization}, author = {{Amir Feder} and {Yoav Wald} and {Claudia Shi} and {S. Saria} and {David Blei}}, year = 2023, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/dce7db550d6edda246d7848a37f777ba3b9bbf2f}, } - Lingfeng Shen, Aayush Mishra, and Daniel Khashabi, “Do pretrained Transformers Really Learn In-context by Gradient Descent?,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263909464, title = {Do pretrained Transformers Really Learn In-context by Gradient Descent?}, author = {{Lingfeng Shen} and {Aayush Mishra} and {Daniel Khashabi}}, year = 2023, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/a9d460f8eb9001b1bed11b7fb2af555185c70fcf}, } - Shiyu Luo, Miguel Angrick, Christopher Coogan, Daniel Candrea, Kimberley Wyse-Sookoo, Samyak Shah, Qinwan Rabbani, Griffin W. Milsap, Alexander R Weiss, William S Anderson, Donna C Tippett, Nicholas J Maragakis, Lora Clawson, M. Vansteensel, B. Wester, Francesco V Tenore, H. Hermansky, M. Fifer, Nick F Ramsey, and Nathan Crone, “Stable Decoding from a Speech BCI Enables Control for an Individual with ALS without Recalibration for 3 Months.,” in Advancement of science, 2023.
[BibTeX] [Link]@inproceedings{264448311, title = {Stable Decoding from a Speech BCI Enables Control for an Individual with ALS without Recalibration for 3 Months.}, author = {{Shiyu Luo} and {Miguel Angrick} and {Christopher Coogan} and {Daniel Candrea} and {Kimberley Wyse-Sookoo} and {Samyak Shah} and {Qinwan Rabbani} and {Griffin W. Milsap} and {Alexander R Weiss} and {William S Anderson} and {Donna C Tippett} and {Nicholas J Maragakis} and {Lora Clawson} and {M. Vansteensel} and {B. Wester} and {Francesco V Tenore} and {H. Hermansky} and {M. Fifer} and {Nick F Ramsey} and {Nathan Crone}}, year = 2023, month = {10}, booktitle = {Advancement of science}, url = {https://www.semanticscholar.org/paper/dee851d6c5652ee423118132e1483bc0af9f30fc}, } - Geoffrey M Gray, Ayah Zirikly, Luis M Ahumada, Masoud Rouhizadeh, Thomas Richards, C. Kitchen, Iman Foroughmand, and E. Hatef, “Application of natural language processing to identify social needs from patient medical notes: development and assessment of a scalable, performant, and rule-based model in an integrated healthcare delivery system,” in JAMIA Open, 2023.
[BibTeX] [Link]@inproceedings{263669151, title = {Application of natural language processing to identify social needs from patient medical notes: development and assessment of a scalable, performant, and rule-based model in an integrated healthcare delivery system}, author = {{Geoffrey M Gray} and {Ayah Zirikly} and {Luis M Ahumada} and {Masoud Rouhizadeh} and {Thomas Richards} and {C. Kitchen} and {Iman Foroughmand} and {E. Hatef}}, year = 2023, month = {10}, booktitle = {JAMIA Open}, url = {https://www.semanticscholar.org/paper/1be931a9ebfeaa018e47abc582b1a9760ced4710}, } - Jiarui Hai, Helin Wang, Dongchao Yang, Karan Thakkar, N. Dehak, and Mounya Elhilali, “DPM-TSE: A Diffusion Probabilistic Model for Target Sound Extraction,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263830793, title = {DPM-TSE: A Diffusion Probabilistic Model for Target Sound Extraction}, author = {{Jiarui Hai} and {Helin Wang} and {Dongchao Yang} and {Karan Thakkar} and {N. Dehak} and {Mounya Elhilali}}, year = 2023, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/70aec6486668cc5ca25d45240c68de223a8deda7}, } - Alex Walker, Ayah Zirikly, Melissa D. Stockbridge, and Holly C Wilcox, “A Linguistic Analysis of Instagram Captions Between Adolescent Suicide Decedents and Living Controls.,” in Crisis, 2023.
[BibTeX] [Link]@inproceedings{263827534, title = {A Linguistic Analysis of Instagram Captions Between Adolescent Suicide Decedents and Living Controls.}, author = {{Alex Walker} and {Ayah Zirikly} and {Melissa D. Stockbridge} and {Holly C Wilcox}}, year = 2023, month = {10}, booktitle = {Crisis}, url = {https://www.semanticscholar.org/paper/a0ee01acead1ccb6064f603f75186f8aa25d2562}, } - Tianjian Li, Haoran Xu, Philipp Koehn, Daniel Khashabi, and Kenton Murray, “Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263605981, title = {Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Models}, author = {{Tianjian Li} and {Haoran Xu} and {Philipp Koehn} and {Daniel Khashabi} and {Kenton Murray}}, year = 2023, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d15021d750dbe9cee120b562acea857ca02d9104}, } - Jiarui Hai and Mounya Elhilali, “Diff-Pitcher: Diffusion-Based Singing Voice Pitch Correction,” in IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, 2023.
[BibTeX] [Link]@inproceedings{261899586, title = {Diff-Pitcher: Diffusion-Based Singing Voice Pitch Correction}, author = {{Jiarui Hai} and {Mounya Elhilali}}, year = 2023, month = {10}, booktitle = {IEEE Workshop on Applications of Signal Processing to Audio and Acoustics}, url = {https://www.semanticscholar.org/paper/377ffdc7cf16822e8aa12ea28ab16d0f5bc8f0c2}, } - Abe Bohan Hou, Jingyu Zhang, Tianxing He, Yichen Wang, Yung-Sung Chuang, Hongwei Wang, Lingfeng Shen, Benjamin Van Durme, Daniel Khashabi, and Yulia Tsvetkov, “SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263831179, title = {SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation}, author = {{Abe Bohan Hou} and {Jingyu Zhang} and {Tianxing He} and {Yichen Wang} and {Yung-Sung Chuang} and {Hongwei Wang} and {Lingfeng Shen} and {Benjamin Van Durme} and {Daniel Khashabi} and {Yulia Tsvetkov}}, year = 2023, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/289533e16e509d0ba8f499a371b4e470f3e492de}, } - A. DeLucia, M. Dredze, and A. L. Buczak, “A Multi-instance Learning Approach to Civil Unrest Event Detection on Twitter,” in Proceedings of the 6th Workshop on Challenges and Applications of Automated Extraction of Socio-political Events from Text, Varna, Bulgaria, 2023, p. 18–33.
[BibTeX] [Abstract] [Link]
Social media has become an established platform for people to organize and take offline actions, often in the form of civil unrest. Understanding these events can help support pro-democratic movements. The primary method to detect these events on Twitter relies on aggregating many tweets, but this includes many that are not relevant to the task. We propose a multi-instance learning (MIL) approach, which jointly identifies relevant tweets and detects civil unrest events. We demonstrate that MIL improves civil unrest detection over methods based on simple aggregation. Our best model achieves a 0.73 F1 on the Global Civil Unrest on Twitter (G-CUT) dataset.
@inproceedings{delucia-etal-2023-multi, title = "A Multi-instance Learning Approach to Civil Unrest Event Detection on {T}witter", author = "DeLucia, Alexandra and Dredze, Mark and Buczak, Anna L.", editor = {H{\"u}rriyeto{\u{g}}lu, Ali and Tanev, Hristo and Zavarella, Vanni and Yeniterzi, Reyyan and Y{\"o}r{\"u}k, Erdem and Slavcheva, Milena}, booktitle = "Proceedings of the 6th Workshop on Challenges and Applications of Automated Extraction of Socio-political Events from Text", month = sep, year = "2023", address = "Varna, Bulgaria", publisher = "INCOMA Ltd., Shoumen, Bulgaria", url = "https://aclanthology.org/2023.case-1.3", pages = "18--33", abstract = "Social media has become an established platform for people to organize and take offline actions, often in the form of civil unrest. Understanding these events can help support pro-democratic movements. The primary method to detect these events on Twitter relies on aggregating many tweets, but this includes many that are not relevant to the task. We propose a multi-instance learning (MIL) approach, which jointly identifies relevant tweets and detects civil unrest events. We demonstrate that MIL improves civil unrest detection over methods based on simple aggregation. Our best model achieves a 0.73 F1 on the Global Civil Unrest on Twitter (G-CUT) dataset.", } - J. Chi, B. Lu, J. Eisner, P. Bell, P. Jyothi, and A. M. Ali, “Unsupervised Code-Switched Text Generation from Parallel Text,” in Proceedings of INTERSPEECH, Dublin, 2023.
[BibTeX] [Link]@InProceedings{chi-et-al-2023, author = "Jie Chi and Brian Lu and Jason Eisner and Peter Bell and Preethi Jyothi and Ahmed M. Ali", title = "Unsupervised Code-Switched Text Generation from Parallel Text", booktitle = "Proceedings of INTERSPEECH", year = "2023", month = aug, address = "Dublin", URL = "http://cs.jhu.edu/~jason/papers/#chi-et-al-2023", } - E. Spaulding, G. Kazantsev, and M. Dredze, “Joint End-to-end Semantic Proto-role Labeling,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Toronto, Canada, 2023, p. 723–736. doi:10.18653/v1/2023.acl-short.63
[BibTeX] [Abstract] [Link]
Semantic proto-role labeling (SPRL) assigns properties to arguments based on a series of binary labels. While multiple studies have evaluated various approaches to SPRL, it has only been studied in-depth as a standalone task using gold predicate/argument pairs. How do SPRL systems perform as part of an information extraction pipeline? We model SPRL jointly with predicate-argument extraction using a deep transformer model. We find that proto-role labeling is surprisingly robust in this setting, with only a small decrease when using predicted arguments. We include a detailed analysis of each component of the joint system, and an error analysis to understand correlations in errors between system stages. Finally, we study the effects of annotation errors on SPRL.
@inproceedings{spaulding-etal-2023-joint, title = "Joint End-to-end Semantic Proto-role Labeling", author = "Spaulding, Elizabeth and Kazantsev, Gary and Dredze, Mark", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-short.63", doi = "10.18653/v1/2023.acl-short.63", pages = "723--736", abstract = "Semantic proto-role labeling (SPRL) assigns properties to arguments based on a series of binary labels. While multiple studies have evaluated various approaches to SPRL, it has only been studied in-depth as a standalone task using gold predicate/argument pairs. How do SPRL systems perform as part of an information extraction pipeline? We model SPRL jointly with predicate-argument extraction using a deep transformer model. We find that proto-role labeling is surprisingly robust in this setting, with only a small decrease when using predicted arguments. We include a detailed analysis of each component of the joint system, and an error analysis to understand correlations in errors between system stages. Finally, we study the effects of annotation errors on SPRL.", } - K. Harrigian, A. Zirikly, B. Chee, A. Ahmad, A. Links, S. Saha, M. C. Beach, and M. Dredze, “Characterization of Stigmatizing Language in Medical Records,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Toronto, Canada, 2023, p. 312–329. doi:10.18653/v1/2023.acl-short.28
[BibTeX] [Abstract] [Link]
Widespread disparities in clinical outcomes exist between different demographic groups in the United States. A new line of work in medical sociology has demonstrated physicians often use stigmatizing language in electronic medical records within certain groups, such as black patients, which may exacerbate disparities. In this study, we characterize these instances at scale using a series of domain-informed NLP techniques. We highlight important differences between this task and analogous bias-related tasks studied within the NLP community (e.g., classifying microaggressions). Our study establishes a foundation for NLP researchers to contribute timely insights to a problem domain brought to the forefront by recent legislation regarding clinical documentation transparency. We release data, code, and models.
@inproceedings{harrigian-etal-2023-characterization, title = "Characterization of Stigmatizing Language in Medical Records", author = "Harrigian, Keith and Zirikly, Ayah and Chee, Brant and Ahmad, Alya and Links, Anne and Saha, Somnath and Beach, Mary Catherine and Dredze, Mark", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-short.28", doi = "10.18653/v1/2023.acl-short.28", pages = "312--329", abstract = "Widespread disparities in clinical outcomes exist between different demographic groups in the United States. A new line of work in medical sociology has demonstrated physicians often use stigmatizing language in electronic medical records within certain groups, such as black patients, which may exacerbate disparities. In this study, we characterize these instances at scale using a series of domain-informed NLP techniques. We highlight important differences between this task and analogous bias-related tasks studied within the NLP community (e.g., classifying microaggressions). Our study establishes a foundation for NLP researchers to contribute timely insights to a problem domain brought to the forefront by recent legislation regarding clinical documentation transparency. We release data, code, and models.", } - S. Zhang, S. Wu, O. Irsoy, S. Lu, M. Bansal, M. Dredze, and D. Rosenberg, “MixCE: Training Autoregressive Language Models by Mixing Forward and Reverse Cross-Entropies,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 9027–9050. doi:10.18653/v1/2023.acl-long.502
[BibTeX] [Abstract] [Link]
Autoregressive language models are trained by minimizing the cross-entropy of the model distribution Q relative to the data distribution P {–} that is, minimizing the forward cross-entropy, which is equivalent to maximum likelihood estimation (MLE). We have observed that models trained in this way may {“}over-generalize{”}, in the sense that they produce non-human-like text. Moreover, we believe that reverse cross-entropy, i.e., the cross-entropy of P relative to Q, is a better reflection of how a human would evaluate text generated by a model. Hence, we propose learning with MixCE, an objective that mixes the forward and reverse cross-entropies. We evaluate models trained with this objective on synthetic data settings (where P is known) and real data, and show that the resulting models yield better generated text without complex decoding strategies.
@inproceedings{zhang-etal-2023-mixce, title = "{M}ix{CE}: Training Autoregressive Language Models by Mixing Forward and Reverse Cross-Entropies", author = "Zhang, Shiyue and Wu, Shijie and Irsoy, Ozan and Lu, Steven and Bansal, Mohit and Dredze, Mark and Rosenberg, David", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.502", doi = "10.18653/v1/2023.acl-long.502", pages = "9027--9050", abstract = "Autoregressive language models are trained by minimizing the cross-entropy of the model distribution Q relative to the data distribution P {--} that is, minimizing the forward cross-entropy, which is equivalent to maximum likelihood estimation (MLE). We have observed that models trained in this way may {``}over-generalize{''}, in the sense that they produce non-human-like text. Moreover, we believe that reverse cross-entropy, i.e., the cross-entropy of P relative to Q, is a better reflection of how a human would evaluate text generated by a model. Hence, we propose learning with MixCE, an objective that mixes the forward and reverse cross-entropies. We evaluate models trained with this objective on synthetic data settings (where P is known) and real data, and show that the resulting models yield better generated text without complex decoding strategies.", } - S. Behzad, S. Ebner, M. Marone, B. Van Durme, and M. Yarmohammadi, “The Effect of Alignment Correction on Cross-Lingual Annotation Projection,” in Proceedings of the 17th Linguistic Annotation Workshop (LAW-XVII), Toronto, Canada, 2023, p. 244–251. doi:10.18653/v1/2023.law-1.24
[BibTeX] [Abstract] [Link]
Cross-lingual annotation projection is a practical method for improving performance on low resource structured prediction tasks. An important step in annotation projection is obtaining alignments between the source and target texts, which enables the mapping of annotations across the texts. By manually correcting automatically generated alignments, we examine the impact of alignment quality{–-}automatic, manual, and mixed{–-}on downstream performance for two information extraction tasks and quantify the trade-off between annotation effort and model performance.
@inproceedings{behzad-etal-2023-effect, title = "The Effect of Alignment Correction on Cross-Lingual Annotation Projection", author = "Behzad, Shabnam and Ebner, Seth and Marone, Marc and Van Durme, Benjamin and Yarmohammadi, Mahsa", editor = "Prange, Jakob and Friedrich, Annemarie", booktitle = "Proceedings of the 17th Linguistic Annotation Workshop (LAW-XVII)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.law-1.24", doi = "10.18653/v1/2023.law-1.24", pages = "244--251", abstract = "Cross-lingual annotation projection is a practical method for improving performance on low resource structured prediction tasks. An important step in annotation projection is obtaining alignments between the source and target texts, which enables the mapping of annotations across the texts. By manually correcting automatically generated alignments, we examine the impact of alignment quality{---}automatic, manual, and mixed{---}on downstream performance for two information extraction tasks and quantify the trade-off between annotation effort and model performance.", } - D. Verma, Y. K. Lal, S. Sinha, B. Van Durme, and A. Poliak, “Evaluating Paraphrastic Robustness in Textual Entailment Models,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Toronto, Canada, 2023, p. 880–892. doi:10.18653/v1/2023.acl-short.76
[BibTeX] [Abstract] [Link]
We present PaRTE, a collection of 1,126 pairs of Recognizing Textual Entailment (RTE) examples to evaluate whether models are robust to paraphrasing. We posit that if RTE models understand language, their predictions should be consistent across inputs that share the same meaning. We use the evaluation set to determine if RTE models{‘} predictions change when examples are paraphrased. In our experiments, contemporary models change their predictions on 8-16{\%} of paraphrased examples, indicating that there is still room for improvement.
@inproceedings{verma-etal-2023-evaluating, title = "Evaluating Paraphrastic Robustness in Textual Entailment Models", author = "Verma, Dhruv and Lal, Yash Kumar and Sinha, Shreyashee and Van Durme, Benjamin and Poliak, Adam", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-short.76", doi = "10.18653/v1/2023.acl-short.76", pages = "880--892", abstract = "We present PaRTE, a collection of 1,126 pairs of Recognizing Textual Entailment (RTE) examples to evaluate whether models are robust to paraphrasing. We posit that if RTE models understand language, their predictions should be consistent across inputs that share the same meaning. We use the evaluation set to determine if RTE models{'} predictions change when examples are paraphrased. In our experiments, contemporary models change their predictions on 8-16{\%} of paraphrased examples, indicating that there is still room for improvement.", } - X. Zhang, K. Duh, and P. McNamee, “A Hyperparameter Optimization Toolkit for Neural Machine Translation Research,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), Toronto, Canada, 2023, p. 161–168. doi:10.18653/v1/2023.acl-demo.15
[BibTeX] [Abstract] [Link]
Hyperparameter optimization is an important but often overlooked process in the research of deep learning technologies. To obtain a good model, one must carefully tune hyperparameters that determine the architecture and training algorithm. Insufficient tuning may result in poor results, while inequitable tuning may lead to exaggerated differences between models. We present a hyperparameter optimization toolkit for neural machine translation (NMT) to help researchers focus their time on the creative rather than the mundane. The toolkit is implemented as a wrapper on top of the open-source Sockeye NMT software. Using the Asynchronous Successive Halving Algorithm (ASHA), we demonstrate that it is possible to discover near-optimal models under a computational budget with little effort. Code: \url{https://github.com/kevinduh/sockeye-recipes3Video} demo: \url{https://cs.jhu.edu/kevinduh/j/demo.mp4}
@inproceedings{zhang-etal-2023-hyperparameter, title = "A Hyperparameter Optimization Toolkit for Neural Machine Translation Research", author = "Zhang, Xuan and Duh, Kevin and McNamee, Paul", editor = "Bollegala, Danushka and Huang, Ruihong and Ritter, Alan", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-demo.15", doi = "10.18653/v1/2023.acl-demo.15", pages = "161--168", abstract = "Hyperparameter optimization is an important but often overlooked process in the research of deep learning technologies. To obtain a good model, one must carefully tune hyperparameters that determine the architecture and training algorithm. Insufficient tuning may result in poor results, while inequitable tuning may lead to exaggerated differences between models. We present a hyperparameter optimization toolkit for neural machine translation (NMT) to help researchers focus their time on the creative rather than the mundane. The toolkit is implemented as a wrapper on top of the open-source Sockeye NMT software. Using the Asynchronous Successive Halving Algorithm (ASHA), we demonstrate that it is possible to discover near-optimal models under a computational budget with little effort. Code: \url{https://github.com/kevinduh/sockeye-recipes3Video} demo: \url{https://cs.jhu.edu/kevinduh/j/demo.mp4}", } - G. Portillo Wightman, A. Delucia, and M. Dredze, “Strength in Numbers: Estimating Confidence of Large Language Models by Prompt Agreement,” in Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023), Toronto, Canada, 2023, p. 326–362. doi:10.18653/v1/2023.trustnlp-1.28
[BibTeX] [Abstract] [Link]
Large language models have achieved impressive few-shot performance on a wide variety of tasks. However, in many settings, users require confidence estimates for model predictions. While traditional classifiers produce scores for each label, language models instead produce scores for the generation which may not be well calibrated. We compare generations across diverse prompts and show that these can be used to create confidence scores. By utilizing more prompts we can get more precise confidence estimates and use response diversity as a proxy for confidence. We evaluate this approach across ten multiple-choice question-answering datasets using three models: T0, FLAN-T5, and GPT-3. In addition to analyzing multiple human written prompts, we automatically generate more prompts using a language model in order to produce finer-grained confidence estimates. Our method produces more calibrated confidence estimates compared to the log probability of the answer to a single prompt. These improvements could benefit users who rely on prediction confidence for integration into a larger system or in decision-making processes.
@inproceedings{portillo-wightman-etal-2023-strength, title = "Strength in Numbers: Estimating Confidence of Large Language Models by Prompt Agreement", author = "Portillo Wightman, Gwenyth and Delucia, Alexandra and Dredze, Mark", editor = "Ovalle, Anaelia and Chang, Kai-Wei and Mehrabi, Ninareh and Pruksachatkun, Yada and Galystan, Aram and Dhamala, Jwala and Verma, Apurv and Cao, Trista and Kumar, Anoop and Gupta, Rahul", booktitle = "Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.trustnlp-1.28", doi = "10.18653/v1/2023.trustnlp-1.28", pages = "326--362", abstract = "Large language models have achieved impressive few-shot performance on a wide variety of tasks. However, in many settings, users require confidence estimates for model predictions. While traditional classifiers produce scores for each label, language models instead produce scores for the generation which may not be well calibrated. We compare generations across diverse prompts and show that these can be used to create confidence scores. By utilizing more prompts we can get more precise confidence estimates and use response diversity as a proxy for confidence. We evaluate this approach across ten multiple-choice question-answering datasets using three models: T0, FLAN-T5, and GPT-3. In addition to analyzing multiple human written prompts, we automatically generate more prompts using a language model in order to produce finer-grained confidence estimates. Our method produces more calibrated confidence estimates compared to the log probability of the answer to a single prompt. These improvements could benefit users who rely on prediction confidence for integration into a larger system or in decision-making processes.", } - J. Gwinnup, T. Anderson, B. Ore, E. Hansen, and K. Duh, “Enhancing Video Translation Context with Object Labels,” in Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023), Toronto, Canada (in-person and online), 2023, p. 130–137. doi:10.18653/v1/2023.iwslt-1.8
[BibTeX] [Abstract] [Link]
We present a simple yet efficient method to enhance the quality of machine translation models trained on multimodal corpora by augmenting the training text with labels of detected objects in the corresponding video segments. We then test the effects of label augmentation in both baseline and two automatic speech recognition (ASR) conditions. In contrast with multimodal techniques that merge visual and textual features, our modular method is easy to implement and the results are more interpretable. Comparisons are made with Transformer translation architectures trained with baseline and augmented labels, showing improvements of up to +1.0 BLEU on the How2 dataset.
@inproceedings{gwinnup-etal-2023-enhancing, title = "Enhancing Video Translation Context with Object Labels", author = "Gwinnup, Jeremy and Anderson, Tim and Ore, Brian and Hansen, Eric and Duh, Kevin", editor = "Salesky, Elizabeth and Federico, Marcello and Carpuat, Marine", booktitle = "Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023)", month = jul, year = "2023", address = "Toronto, Canada (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.iwslt-1.8", doi = "10.18653/v1/2023.iwslt-1.8", pages = "130--137", abstract = "We present a simple yet efficient method to enhance the quality of machine translation models trained on multimodal corpora by augmenting the training text with labels of detected objects in the corresponding video segments. We then test the effects of label augmentation in both baseline and two automatic speech recognition (ASR) conditions. In contrast with multimodal techniques that merge visual and textual features, our modular method is easy to implement and the results are more interpretable. Comparisons are made with Transformer translation architectures trained with baseline and augmented labels, showing improvements of up to +1.0 BLEU on the How2 dataset.", } - M. Agarwal, S. Agrawal, A. Anastasopoulos, L. Bentivogli, O. Bojar, C. Borg, M. Carpuat, R. Cattoni, M. Cettolo, M. Chen, W. Chen, K. Choukri, A. Chronopoulou, A. Currey, T. Declerck, Q. Dong, K. Duh, Y. Estève, M. Federico, S. Gahbiche, B. Haddow, B. Hsu, P. Mon Htut, H. Inaguma, D. Javorský, J. Judge, Y. Kano, T. Ko, R. Kumar, P. Li, X. Ma, P. Mathur, E. Matusov, P. McNamee, J. P. McCrae, K. Murray, M. Nadejde, S. Nakamura, M. Negri, H. Nguyen, J. Niehues, X. Niu, A. Kr. Ojha, J. E. Ortega, P. Pal, J. Pino, L. van der Plas, P. Polák, E. Rippeth, E. Salesky, J. Shi, M. Sperber, S. Stüker, K. Sudoh, Y. Tang, B. Thompson, K. Tran, M. Turchi, A. Waibel, M. Wang, S. Watanabe, and R. Zevallos, “FINDINGS OF THE IWSLT 2023 EVALUATION CAMPAIGN,” in Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023), Toronto, Canada (in-person and online), 2023, p. 1–61. doi:10.18653/v1/2023.iwslt-1.1
[BibTeX] [Abstract] [Link]
This paper reports on the shared tasks organized by the 20th IWSLT Conference. The shared tasks address 9 scientific challenges in spoken language translation: simultaneous and offline translation, automatic subtitling and dubbing, speech-to-speech translation, multilingual, dialect and low-resource speech translation, and formality control. The shared tasks attracted a total of 38 submissions by 31 teams. The growing interest towards spoken language translation is also witnessed by the constantly increasing number of shared task organizers and contributors to the overview paper, almost evenly distributed across industry and academia.
@inproceedings{agrawal-etal-2023-findings, title = "{FINDINGS} {OF} {THE} {IWSLT} 2023 {EVALUATION} {CAMPAIGN}", author = {Agarwal, Milind and Agrawal, Sweta and Anastasopoulos, Antonios and Bentivogli, Luisa and Bojar, Ond{\v{r}}ej and Borg, Claudia and Carpuat, Marine and Cattoni, Roldano and Cettolo, Mauro and Chen, Mingda and Chen, William and Choukri, Khalid and Chronopoulou, Alexandra and Currey, Anna and Declerck, Thierry and Dong, Qianqian and Duh, Kevin and Est{\`e}ve, Yannick and Federico, Marcello and Gahbiche, Souhir and Haddow, Barry and Hsu, Benjamin and Mon Htut, Phu and Inaguma, Hirofumi and Javorsk{\'y}, D{\'a}vid and Judge, John and Kano, Yasumasa and Ko, Tom and Kumar, Rishu and Li, Pengwei and Ma, Xutai and Mathur, Prashant and Matusov, Evgeny and McNamee, Paul and P. McCrae, John and Murray, Kenton and Nadejde, Maria and Nakamura, Satoshi and Negri, Matteo and Nguyen, Ha and Niehues, Jan and Niu, Xing and Kr. Ojha, Atul and E. Ortega, John and Pal, Proyag and Pino, Juan and van der Plas, Lonneke and Pol{\'a}k, Peter and Rippeth, Elijah and Salesky, Elizabeth and Shi, Jiatong and Sperber, Matthias and St{\"u}ker, Sebastian and Sudoh, Katsuhito and Tang, Yun and Thompson, Brian and Tran, Kevin and Turchi, Marco and Waibel, Alex and Wang, Mingxuan and Watanabe, Shinji and Zevallos, Rodolfo}, editor = "Salesky, Elizabeth and Federico, Marcello and Carpuat, Marine", booktitle = "Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023)", month = jul, year = "2023", address = "Toronto, Canada (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.iwslt-1.1", doi = "10.18653/v1/2023.iwslt-1.1", pages = "1--61", abstract = "This paper reports on the shared tasks organized by the 20th IWSLT Conference. The shared tasks address 9 scientific challenges in spoken language translation: simultaneous and offline translation, automatic subtitling and dubbing, speech-to-speech translation, multilingual, dialect and low-resource speech translation, and formality control. The shared tasks attracted a total of 38 submissions by 31 teams. The growing interest towards spoken language translation is also witnessed by the constantly increasing number of shared task organizers and contributors to the overview paper, almost evenly distributed across industry and academia.", } - A. Hussein, C. Xiao, N. Verma, T. Thebaud, M. Wiesner, and S. Khudanpur, “JHU IWSLT 2023 Dialect Speech Translation System Description,” in Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023), Toronto, Canada (in-person and online), 2023, p. 283–290. doi:10.18653/v1/2023.iwslt-1.26
[BibTeX] [Abstract] [Link]
This paper presents JHU{‘}s submissions to the IWSLT 2023 dialectal and low-resource track of Tunisian Arabic to English speech translation. The Tunisian dialect lacks formal orthography and abundant training data, making it challenging to develop effective speech translation (ST) systems. To address these challenges, we explore the integration of large pre-trained machine translation (MT) models, such as mBART and NLLB-200 in both end-to-end (E2E) and cascaded speech translation (ST) systems. We also improve the performance of automatic speech recognition (ASR) through the use of pseudo-labeling data augmentation and channel matching on telephone data. Finally, we combine our E2E and cascaded ST systems with Minimum Bayes-Risk decoding. Our combined system achieves a BLEU score of 21.6 and 19.1 on test2 and test3, respectively.
@inproceedings{hussein-etal-2023-jhu, title = "{JHU} {IWSLT} 2023 Dialect Speech Translation System Description", author = "Hussein, Amir and Xiao, Cihan and Verma, Neha and Thebaud, Thomas and Wiesner, Matthew and Khudanpur, Sanjeev", editor = "Salesky, Elizabeth and Federico, Marcello and Carpuat, Marine", booktitle = "Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023)", month = jul, year = "2023", address = "Toronto, Canada (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.iwslt-1.26", doi = "10.18653/v1/2023.iwslt-1.26", pages = "283--290", abstract = "This paper presents JHU{'}s submissions to the IWSLT 2023 dialectal and low-resource track of Tunisian Arabic to English speech translation. The Tunisian dialect lacks formal orthography and abundant training data, making it challenging to develop effective speech translation (ST) systems. To address these challenges, we explore the integration of large pre-trained machine translation (MT) models, such as mBART and NLLB-200 in both end-to-end (E2E) and cascaded speech translation (ST) systems. We also improve the performance of automatic speech recognition (ASR) through the use of pseudo-labeling data augmentation and channel matching on telephone data. Finally, we combine our E2E and cascaded ST systems with Minimum Bayes-Risk decoding. Our combined system achieves a BLEU score of 21.6 and 19.1 on test2 and test3, respectively.", } - H. L. Xinyuan, N. Verma, B. Bamfo Odoom, U. Pradeep, M. Wiesner, and S. Khudanpur, “JHU IWSLT 2023 Multilingual Speech Translation System Description,” in Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023), Toronto, Canada (in-person and online), 2023, p. 302–310. doi:10.18653/v1/2023.iwslt-1.28
[BibTeX] [Abstract] [Link]
We describe the Johns Hopkins ACL 60-60 Speech Translation systems submitted to the IWSLT 2023 Multilingual track, where we were tasked to translate ACL presentations from English into 10 languages. We developed cascaded speech translation systems for both the constrained and unconstrained subtracks. Our systems make use of pre-trained models as well as domain-specific corpora for this highly technical evaluation-only task. We find that the specific technical domain which ACL presentations fall into presents a unique challenge for both ASR and MT, and we present an error analysis and an ACL-specific corpus we produced to enable further work in this area.
@inproceedings{xinyuan-etal-2023-jhu, title = "{JHU} {IWSLT} 2023 Multilingual Speech Translation System Description", author = "Xinyuan, Henry Li and Verma, Neha and Bamfo Odoom, Bismarck and Pradeep, Ujvala and Wiesner, Matthew and Khudanpur, Sanjeev", editor = "Salesky, Elizabeth and Federico, Marcello and Carpuat, Marine", booktitle = "Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023)", month = jul, year = "2023", address = "Toronto, Canada (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.iwslt-1.28", doi = "10.18653/v1/2023.iwslt-1.28", pages = "302--310", abstract = "We describe the Johns Hopkins ACL 60-60 Speech Translation systems submitted to the IWSLT 2023 Multilingual track, where we were tasked to translate ACL presentations from English into 10 languages. We developed cascaded speech translation systems for both the constrained and unconstrained subtracks. Our systems make use of pre-trained models as well as domain-specific corpora for this highly technical evaluation-only task. We find that the specific technical domain which ACL presentations fall into presents a unique challenge for both ASR and MT, and we present an error analysis and an ACL-specific corpus we produced to enable further work in this area.", } - J. Maillard, C. Gao, E. Kalbassi, K. R. Sadagopan, V. Goswami, P. Koehn, A. Fan, and F. Guzman, “Small Data, Big Impact: Leveraging Minimal Data for Effective Machine Translation,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 2740–2756. doi:10.18653/v1/2023.acl-long.154
[BibTeX] [Abstract] [Link]
For many languages, machine translation progress is hindered by the lack of reliable training data. Models are trained on whatever pre-existing datasets may be available and then augmented with synthetic data, because it is often not economical to pay for the creation of large-scale datasets. But for the case of low-resource languages, would the creation of a few thousand professionally translated sentence pairs give any benefit? In this paper, we show that it does. We describe a broad data collection effort involving around 6k professionally translated sentence pairs for each of 39 low-resource languages, which we make publicly available. We analyse the gains of models trained on this small but high-quality data, showing that it has significant impact even when larger but lower quality pre-existing corpora are used, or when data is augmented with millions of sentences through backtranslation.
@inproceedings{maillard-etal-2023-small, title = "Small Data, Big Impact: Leveraging Minimal Data for Effective Machine Translation", author = "Maillard, Jean and Gao, Cynthia and Kalbassi, Elahe and Sadagopan, Kaushik Ram and Goswami, Vedanuj and Koehn, Philipp and Fan, Angela and Guzman, Francisco", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.154", doi = "10.18653/v1/2023.acl-long.154", pages = "2740--2756", abstract = "For many languages, machine translation progress is hindered by the lack of reliable training data. Models are trained on whatever pre-existing datasets may be available and then augmented with synthetic data, because it is often not economical to pay for the creation of large-scale datasets. But for the case of low-resource languages, would the creation of a few thousand professionally translated sentence pairs give any benefit? In this paper, we show that it does. We describe a broad data collection effort involving around 6k professionally translated sentence pairs for each of 39 low-resource languages, which we make publicly available. We analyse the gains of models trained on this small but high-quality data, showing that it has significant impact even when larger but lower quality pre-existing corpora are used, or when data is augmented with millions of sentences through backtranslation.", } - E. Schumacher, J. Mayfield, and M. Dredze, “On the Surprising Effectiveness of Name Matching Alone in Autoregressive Entity Linking,” in Proceedings of the First Workshop on Matching From Unstructured and Structured Data (MATCHING 2023), Toronto, ON, Canada, 2023, p. 58–69. doi:10.18653/v1/2023.matching-1.6
[BibTeX] [Abstract] [Link]
Fifteen years of work on entity linking has established the importance of different information sources in making linking decisions: mention and entity name similarity, contextual relevance, and features of the knowledge base. Modern state-of-the-art systems build on these features, including through neural representations (Wu et al., 2020). In contrast to this trend, the autoregressive language model GENRE (De Cao et al., 2021) generates normalized entity names for mentions and beats many other entity linking systems, despite making no use of knowledge base (KB) information. How is this possible? We analyze the behavior of GENRE on several entity linking datasets and demonstrate that its performance stems from memorization of name patterns. In contrast, it fails in cases that might benefit from using the KB. We experiment with a modification to the model to enable it to utilize KB information, highlighting challenges to incorporating traditional entity linking information sources into autoregressive models.
@inproceedings{schumacher-etal-2023-surprising, title = "On the Surprising Effectiveness of Name Matching Alone in Autoregressive Entity Linking", author = "Schumacher, Elliot and Mayfield, James and Dredze, Mark", editor = "Hruschka, Estevam and Mitchell, Tom and Rahman, Sajjadur and Mladeni{\'c}, Dunja and Grobelnik, Marko", booktitle = "Proceedings of the First Workshop on Matching From Unstructured and Structured Data (MATCHING 2023)", month = jul, year = "2023", address = "Toronto, ON, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.matching-1.6", doi = "10.18653/v1/2023.matching-1.6", pages = "58--69", abstract = "Fifteen years of work on entity linking has established the importance of different information sources in making linking decisions: mention and entity name similarity, contextual relevance, and features of the knowledge base. Modern state-of-the-art systems build on these features, including through neural representations (Wu et al., 2020). In contrast to this trend, the autoregressive language model GENRE (De Cao et al., 2021) generates normalized entity names for mentions and beats many other entity linking systems, despite making no use of knowledge base (KB) information. How is this possible? We analyze the behavior of GENRE on several entity linking datasets and demonstrate that its performance stems from memorization of name patterns. In contrast, it fails in cases that might benefit from using the KB. We experiment with a modification to the model to enable it to utilize KB information, highlighting challenges to incorporating traditional entity linking information sources into autoregressive models.", } - M. Antoniak, A. Field, J. Mun, M. Walsh, L. Klein, and M. Sap, “Riveter: Measuring Power and Social Dynamics Between Entities,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), Toronto, Canada, 2023, p. 377–388. doi:10.18653/v1/2023.acl-demo.36
[BibTeX] [Abstract] [Link]
Riveter provides a complete easy-to-use pipeline for analyzing verb connotations associated with entities in text corpora. We prepopulate the package with connotation frames of sentiment, power, and agency, which have demonstrated usefulness for capturing social phenomena, such as gender bias, in a broad range of corpora. For decades, lexical frameworks have been foundational tools in computational social science, digital humanities, and natural language processing, facilitating multifaceted analysis of text corpora. But working with verb-centric lexica specifically requires natural language processing skills, reducing their accessibility to other researchers. By organizing the language processing pipeline, providing complete lexicon scores and visualizations for all entities in a corpus, and providing functionality for users to target specific research questions, Riveter greatly improves the accessibility of verb lexica and can facilitate a broad range of future research.
@inproceedings{antoniak-etal-2023-riveter, title = "Riveter: Measuring Power and Social Dynamics Between Entities", author = "Antoniak, Maria and Field, Anjalie and Mun, Jimin and Walsh, Melanie and Klein, Lauren and Sap, Maarten", editor = "Bollegala, Danushka and Huang, Ruihong and Ritter, Alan", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-demo.36", doi = "10.18653/v1/2023.acl-demo.36", pages = "377--388", abstract = "Riveter provides a complete easy-to-use pipeline for analyzing verb connotations associated with entities in text corpora. We prepopulate the package with connotation frames of sentiment, power, and agency, which have demonstrated usefulness for capturing social phenomena, such as gender bias, in a broad range of corpora. For decades, lexical frameworks have been foundational tools in computational social science, digital humanities, and natural language processing, facilitating multifaceted analysis of text corpora. But working with verb-centric lexica specifically requires natural language processing skills, reducing their accessibility to other researchers. By organizing the language processing pipeline, providing complete lexicon scores and visualizations for all entities in a corpus, and providing functionality for users to target specific research questions, Riveter greatly improves the accessibility of verb lexica and can facilitate a broad range of future research.", } - E. Stengel-Eskin, J. Guallar-Blasco, Y. Zhou, and B. Van Durme, “Why Did the Chicken Cross the Road? Rephrasing and Analyzing Ambiguous Questions in VQA,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 10220–10237. doi:10.18653/v1/2023.acl-long.569
[BibTeX] [Abstract] [Link]
Natural language is ambiguous. Resolving ambiguous questions is key to successfully answering them. Focusing on questions about images, we create a dataset of ambiguous examples. We annotate these, grouping answers by the underlying question they address and rephrasing the question for each group to reduce ambiguity. Our analysis reveals a linguistically-aligned ontology of reasons for ambiguity in visual questions. We then develop an English question-generation model which we demonstrate via automatic and human evaluation produces less ambiguous questions. We further show that the question generation objective we use allows the model to integrate answer group information without any direct supervision.
@inproceedings{stengel-eskin-etal-2023-chicken, title = "Why Did the Chicken Cross the Road? Rephrasing and Analyzing Ambiguous Questions in {VQA}", author = "Stengel-Eskin, Elias and Guallar-Blasco, Jimena and Zhou, Yi and Van Durme, Benjamin", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.569", doi = "10.18653/v1/2023.acl-long.569", pages = "10220--10237", abstract = "Natural language is ambiguous. Resolving ambiguous questions is key to successfully answering them. Focusing on questions about images, we create a dataset of ambiguous examples. We annotate these, grouping answers by the underlying question they address and rephrasing the question for each group to reduce ambiguity. Our analysis reveals a linguistically-aligned ontology of reasons for ambiguity in visual questions. We then develop an English question-generation model which we demonstrate via automatic and human evaluation produces less ambiguous questions. We further show that the question generation objective we use allows the model to integrate answer group information without any direct supervision.", } - A. Mallen, A. Asai, V. Zhong, R. Das, D. Khashabi, and H. Hajishirzi, “When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 9802–9822. doi:10.18653/v1/2023.acl-long.546
[BibTeX] [Abstract] [Link]
Despite their impressive performance on diverse tasks, large language models (LMs) still struggle with tasks requiring rich world knowledge, implying the difficulty of encoding a wealth of world knowledge in their parameters. This paper aims to understand LMs{‘} strengths and limitations in memorizing factual knowledge, by conducting large-scale knowledge probing experiments on two open-domain entity-centric QA datasets: PopQA, our new dataset with 14k questions about long-tail entities, and EntityQuestions, a widely used open-domain QA dataset. We find that LMs struggle with less popular factual knowledge, and that retrieval augmentation helps significantly in these cases. Scaling, on the other hand, mainly improves memorization of popular knowledge, and fails to appreciably improve memorization of factual knowledge in the tail. Based on those findings, we devise a new method for retrieval-augmentation that improves performance and reduces inference costs by only retrieving non-parametric memories when necessary.
@inproceedings{mallen-etal-2023-trust, title = "When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories", author = "Mallen, Alex and Asai, Akari and Zhong, Victor and Das, Rajarshi and Khashabi, Daniel and Hajishirzi, Hannaneh", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.546", doi = "10.18653/v1/2023.acl-long.546", pages = "9802--9822", abstract = "Despite their impressive performance on diverse tasks, large language models (LMs) still struggle with tasks requiring rich world knowledge, implying the difficulty of encoding a wealth of world knowledge in their parameters. This paper aims to understand LMs{'} strengths and limitations in memorizing factual knowledge, by conducting large-scale knowledge probing experiments on two open-domain entity-centric QA datasets: PopQA, our new dataset with 14k questions about long-tail entities, and EntityQuestions, a widely used open-domain QA dataset. We find that LMs struggle with less popular factual knowledge, and that retrieval augmentation helps significantly in these cases. Scaling, on the other hand, mainly improves memorization of popular knowledge, and fails to appreciably improve memorization of factual knowledge in the tail. Based on those findings, we devise a new method for retrieval-augmentation that improves performance and reduces inference costs by only retrieving non-parametric memories when necessary.", } - Y. Wang, Y. Kordi, S. Mishra, A. Liu, N. A. Smith, D. Khashabi, and H. Hajishirzi, “Self-Instruct: Aligning Language Models with Self-Generated Instructions,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 13484–13508. doi:10.18653/v1/2023.acl-long.754
[BibTeX] [Abstract] [Link]
Large {“}instruction-tuned{”} language models (i.e., finetuned to respond to instructions) have demonstrated a remarkable ability to generalize zero-shot to new tasks. Nevertheless, they depend heavily on human-written instruction data that is often limited in quantity, diversity, and creativity, therefore hindering the generality of the tuned model. We introduce Self-Instruct, a framework for improving the instruction-following capabilities of pretrained language models by bootstrapping off their own generations. Our pipeline generates instructions, input, and output samples from a language model, then filters invalid or similar ones before using them to finetune the original model. Applying our method to the vanilla GPT3, we demonstrate a 33{\%} absolute improvement over the original model on Super-NaturalInstructions, on par with the performance of InstructGPT-001, which was trained with private user data and human annotations. For further evaluation, we curate a set of expert-written instructions for novel tasks, and show through human evaluation that tuning GPT3 with Self-Instruct outperforms using existing public instruction datasets by a large margin, leaving only a 5{\%} absolute gap behind InstructGPT-001. Self-Instruct provides an almost annotation-free method for aligning pre-trained language models with instructions, and we release our large synthetic dataset to facilitate future studies on instruction tuning.
@inproceedings{wang-etal-2023-self-instruct, title = "Self-Instruct: Aligning Language Models with Self-Generated Instructions", author = "Wang, Yizhong and Kordi, Yeganeh and Mishra, Swaroop and Liu, Alisa and Smith, Noah A. and Khashabi, Daniel and Hajishirzi, Hannaneh", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.754", doi = "10.18653/v1/2023.acl-long.754", pages = "13484--13508", abstract = "Large {``}instruction-tuned{''} language models (i.e., finetuned to respond to instructions) have demonstrated a remarkable ability to generalize zero-shot to new tasks. Nevertheless, they depend heavily on human-written instruction data that is often limited in quantity, diversity, and creativity, therefore hindering the generality of the tuned model. We introduce Self-Instruct, a framework for improving the instruction-following capabilities of pretrained language models by bootstrapping off their own generations. Our pipeline generates instructions, input, and output samples from a language model, then filters invalid or similar ones before using them to finetune the original model. Applying our method to the vanilla GPT3, we demonstrate a 33{\%} absolute improvement over the original model on Super-NaturalInstructions, on par with the performance of InstructGPT-001, which was trained with private user data and human annotations. For further evaluation, we curate a set of expert-written instructions for novel tasks, and show through human evaluation that tuning GPT3 with Self-Instruct outperforms using existing public instruction datasets by a large margin, leaving only a 5{\%} absolute gap behind InstructGPT-001. Self-Instruct provides an almost annotation-free method for aligning pre-trained language models with instructions, and we release our large synthetic dataset to facilitate future studies on instruction tuning.", } - N. Selvam, S. Dev, D. Khashabi, T. Khot, and K. Chang, “The Tail Wagging the Dog: Dataset Construction Biases of Social Bias Benchmarks,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Toronto, Canada, 2023, p. 1373–1386. doi:10.18653/v1/2023.acl-short.118
[BibTeX] [Abstract] [Link]
How reliably can we trust the scores obtained from social bias benchmarks as faithful indicators of problematic social biases in a given model? In this work, we study this question by contrasting social biases with non-social biases that stem from choices made during dataset construction (which might not even be discernible to the human eye). To do so, we empirically simulate various alternative constructions for a given benchmark based on seemingly innocuous modifications (such as paraphrasing or random-sampling) that maintain the essence of their social bias. On two well-known social bias benchmarks (Winogender and BiasNLI), we observe that these shallow modifications have a surprising effect on the resulting degree of bias across various models and consequently the relative ordering of these models when ranked by measured bias. We hope these troubling observations motivate more robust measures of social biases.
@inproceedings{selvam-etal-2023-tail, title = "The Tail Wagging the Dog: Dataset Construction Biases of Social Bias Benchmarks", author = "Selvam, Nikil and Dev, Sunipa and Khashabi, Daniel and Khot, Tushar and Chang, Kai-Wei", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-short.118", doi = "10.18653/v1/2023.acl-short.118", pages = "1373--1386", abstract = "How reliably can we trust the scores obtained from social bias benchmarks as faithful indicators of problematic social biases in a given model? In this work, we study this question by contrasting social biases with non-social biases that stem from choices made during dataset construction (which might not even be discernible to the human eye). To do so, we empirically simulate various alternative constructions for a given benchmark based on seemingly innocuous modifications (such as paraphrasing or random-sampling) that maintain the essence of their social bias. On two well-known social bias benchmarks (Winogender and BiasNLI), we observe that these shallow modifications have a surprising effect on the resulting degree of bias across various models and consequently the relative ordering of these models when ranked by measured bias. We hope these troubling observations motivate more robust measures of social biases.", } - N. Gandhi, A. Field, and E. Strubell, “Annotating Mentions Alone Enables Efficient Domain Adaptation for Coreference Resolution,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada, 2023, p. 10543–10558. doi:10.18653/v1/2023.acl-long.588
[BibTeX] [Abstract] [Link]
Although recent neural models for coreference resolution have led to substantial improvements on benchmark datasets, it remains a challenge to successfully transfer these models to new target domains containing many out-of-vocabulary spans and requiring differing annotation schemes. Typical approaches involve continued training on annotated target-domain data, but obtaining annotations is costly and time-consuming. In this work, we show that adapting mention detection is the key component to successful domain adaptation of coreference models, rather than antecedent linking. We also show annotating mentions alone is nearly twice as fast as annotating full coreference chains. Based on these insights, we propose a method for efficiently adapting coreference models, which includes a high-precision mention detection objective and requires only mention annotations in the target domain. Extensive evaluation across three English coreference datasets: CoNLL-2012 (news/conversation), i2b2/VA (medical notes), and child welfare notes, reveals that our approach facilitates annotation-efficient transfer and results in a 7-14{\%} improvement in average F1 without increasing annotator time.
@inproceedings{gandhi-etal-2023-annotating, title = "Annotating Mentions Alone Enables Efficient Domain Adaptation for Coreference Resolution", author = "Gandhi, Nupoor and Field, Anjalie and Strubell, Emma", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-long.588", doi = "10.18653/v1/2023.acl-long.588", pages = "10543--10558", abstract = "Although recent neural models for coreference resolution have led to substantial improvements on benchmark datasets, it remains a challenge to successfully transfer these models to new target domains containing many out-of-vocabulary spans and requiring differing annotation schemes. Typical approaches involve continued training on annotated target-domain data, but obtaining annotations is costly and time-consuming. In this work, we show that adapting mention detection is the key component to successful domain adaptation of coreference models, rather than antecedent linking. We also show annotating mentions alone is nearly twice as fast as annotating full coreference chains. Based on these insights, we propose a method for efficiently adapting coreference models, which includes a high-precision mention detection objective and requires only mention annotations in the target domain. Extensive evaluation across three English coreference datasets: CoNLL-2012 (news/conversation), i2b2/VA (medical notes), and child welfare notes, reveals that our approach facilitates annotation-efficient transfer and results in a 7-14{\%} improvement in average F1 without increasing annotator time.", } - H. Fang, A. Balakrishnan, H. Jhamtani, J. Bufe, J. Crawford, Jayant Krishnamurthy, A. Pauls, J. Eisner, Jacob Andreas, and D. Klein, “The Whole Truth and Nothing But the Truth: Faithful and Controllable Dialogue Response Generation with Dataflow Transduction and Constrained Decoding,” in Findings of the Association for Computational Linguistics: ACL 2023, 2023, p. 5682–5700.
[BibTeX] [Link]@InProceedings{fang-et-al-2023, author = "Hao Fang and Anusha Balakrishnan and Harsh Jhamtani and John Bufe and Jean Crawford and Jayant Krishnamurthy and Adam Pauls and Jason Eisner and Jacob Andreas and Dan Klein", title = "The Whole Truth and Nothing But the Truth: Faithful and Controllable Dialogue Response Generation with Dataflow Transduction and Constrained Decoding", booktitle = "Findings of the Association for Computational Linguistics: ACL 2023", year = "2023", month = jul, pages = "5682--5700", URL = "http://cs.jhu.edu/~jason/papers/#fang-et-al-2023", } - B. Z. Li, J. Eisner, A. Pauls, and Sam Thomson, “Toward Interactive Dictation,” in Proceedings of the Association for Computational Linguistics (ACL), 2023, p. 15319–15338.
[BibTeX] [Link]@InProceedings{li-et-al-2023-dictation, author = "Belinda Z. Li and Jason Eisner and Adam Pauls and Sam Thomson", title = "Toward Interactive Dictation", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2023", month = jul, pages = "15319--15338", URL = "http://cs.jhu.edu/~jason/papers/#li-et-al-2023-dictation", } - F. Mireshghallah, Y. Su, Tatsunori Hashimoto, J. Eisner, and R. Shin, “Privacy-Preserving Domain Adaptation of Semantic Parsers,” in Proceedings of the Association for Computational Linguistics (ACL), 2023, p. 4950–4970.
[BibTeX] [Link]@InProceedings{mireshghallah-et-al-2023, author = "Fatemehsadat Mireshghallah and Yu Su and Tatsunori Hashimoto and Jason Eisner and Richard Shin", title = "Privacy-Preserving Domain Adaptation of Semantic Parsers", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2023", month = jul, pages = "4950--4970", URL = "http://cs.jhu.edu/~jason/papers/#mireshghallah-et-al-2023", } - X. L. Li, A. Holtzman, D. Fried, P. Liang, J. Eisner, T. Hashimoto, L. Zettlemoyer, and M. Lewis, “Contrastive Decoding: Open-ended Text Generation as Optimization,” in Proceedings of the Association for Computational Linguistics (ACL), 2023, p. 12286–12312.
[BibTeX] [Link]@InProceedings{li-et-al-2023-cd, author = "Xiang Lisa Li and Ari Holtzman and Daniel Fried and Percy Liang and Jason Eisner and Tatsunori Hashimoto and Luke Zettlemoyer and Mike Lewis", title = "Contrastive Decoding: Open-ended Text Generation as Optimization", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2023", month = jul, pages = "12286--12312", URL = "http://cs.jhu.edu/~jason/papers/#li-et-al-2023-cd", } - L. Du, L. T. Hennigen, T. Pimentel, C. Meister, J. Eisner, and R. Cotterell, “A Measure-Theoretic Characterization of Tight Language Models,” in Proceedings of the Association for Computational Linguistics (ACL), 2023, p. 9744–9770.
[BibTeX] [Link]@InProceedings{du-et-al-2023, author = "Li Du and Lucas Torroba Hennigen and Tiago Pimentel and Clara Meister and Jason Eisner and Ryan Cotterell", title = "A Measure-Theoretic Characterization of Tight Language Models", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2023", month = jul, pages = "9744--9770", URL = "http://cs.jhu.edu/~jason/papers/#du-et-al-2023", } - A. Opedal, R. Zmigrod, T. Vieira, Ryan Cotterell, and J. Eisner, “Efficient Semiring-Weighted Earley Parsing,” in Proceedings of the Association for Computational Linguistics (ACL), 2023, p. 3687–3713.
[BibTeX] [Link]@InProceedings{opedal-et-al-2023, author = "Andreas Opedal and Ran Zmigrod and Tim Vieira and Ryan Cotterell and Jason Eisner", title = "Efficient Semiring-Weighted {E}arley Parsing", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2023", month = jul, pages = "3687--3713", URL = "http://cs.jhu.edu/~jason/papers/#opedal-et-al-2023", } - W. Tan, K. Heffernan, H. Schwenk, and P. Koehn, “Multilingual Representation Distillation with Contrastive Learning,” in Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, Dubrovnik, Croatia, 2023, p. 1477–1490. doi:10.18653/v1/2023.eacl-main.108
[BibTeX] [Abstract] [Link]
Multilingual sentence representations from large models encode semantic information from two or more languages and can be used for different cross-lingual information retrieval and matching tasks. In this paper, we integrate contrastive learning into multilingual representation distillation and use it for quality estimation of parallel sentences (i.e., find semantically similar sentences that can be used as translations of each other). We validate our approach with multilingual similarity search and corpus filtering tasks. Experiments across different low-resource languages show that our method greatly outperforms previous sentence encoders such as LASER, LASER3, and LaBSE.
@inproceedings{tan-etal-2023-multilingual, title = "Multilingual Representation Distillation with Contrastive Learning", author = "Tan, Weiting and Heffernan, Kevin and Schwenk, Holger and Koehn, Philipp", editor = "Vlachos, Andreas and Augenstein, Isabelle", booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics", month = may, year = "2023", address = "Dubrovnik, Croatia", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.eacl-main.108", doi = "10.18653/v1/2023.eacl-main.108", pages = "1477--1490", abstract = "Multilingual sentence representations from large models encode semantic information from two or more languages and can be used for different cross-lingual information retrieval and matching tasks. In this paper, we integrate contrastive learning into multilingual representation distillation and use it for quality estimation of parallel sentences (i.e., find semantically similar sentences that can be used as translations of each other). We validate our approach with multilingual similarity search and corpus filtering tasks. Experiments across different low-resource languages show that our method greatly outperforms previous sentence encoders such as LASER, LASER3, and LaBSE.", } - G. Qin, Y. Feng, and B. Van Durme, “The NLP Task Effectiveness of Long-Range Transformers,” in Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, Dubrovnik, Croatia, 2023, p. 3774–3790. doi:10.18653/v1/2023.eacl-main.273
[BibTeX] [Abstract] [Link]
Transformer models cannot easily scale to long sequences due to their O(N{\^{}}2) time and space complexity. This has led to Transformer variants seeking to lower computational complexity, such as Longformer and Performer. While such models have theoretically greater efficiency, their effectiveness on real NLP tasks has not been well studied. We benchmark 7 variants of Transformer models on 5 difficult NLP tasks and 7 datasets. We design experiments to isolate the effect of pretraining and hyperparameter settings, to focus on their capacity for long-range attention. Moreover, we present various methods to investigate attention behaviors to illuminate model details beyond metric scores. We find that the modified attention in long-range transformers has advantages on content selection and query-guided decoding, but they come with previously unrecognized drawbacks such as insufficient attention to distant tokens and accumulated approximation error.
@inproceedings{qin-etal-2023-nlp, title = "The {NLP} Task Effectiveness of Long-Range Transformers", author = "Qin, Guanghui and Feng, Yukun and Van Durme, Benjamin", editor = "Vlachos, Andreas and Augenstein, Isabelle", booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics", month = may, year = "2023", address = "Dubrovnik, Croatia", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.eacl-main.273", doi = "10.18653/v1/2023.eacl-main.273", pages = "3774--3790", abstract = "Transformer models cannot easily scale to long sequences due to their O(N{\^{}}2) time and space complexity. This has led to Transformer variants seeking to lower computational complexity, such as Longformer and Performer. While such models have theoretically greater efficiency, their effectiveness on real NLP tasks has not been well studied. We benchmark 7 variants of Transformer models on 5 difficult NLP tasks and 7 datasets. We design experiments to isolate the effect of pretraining and hyperparameter settings, to focus on their capacity for long-range attention. Moreover, we present various methods to investigate attention behaviors to illuminate model details beyond metric scores. We find that the modified attention in long-range transformers has advantages on content selection and query-guided decoding, but they come with previously unrecognized drawbacks such as insufficient attention to distant tokens and accumulated approximation error.", } - Y. Chen, W. Gantt, W. Gu, T. Chen, A. White, and B. Van Durme, “Iterative Document-level Information Extraction via Imitation Learning,” in Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, Dubrovnik, Croatia, 2023, p. 1858–1874. doi:10.18653/v1/2023.eacl-main.136
[BibTeX] [Abstract] [Link]
We present a novel iterative extraction model, IterX, for extracting complex relations, or templates, i.e., N-tuples representing a mapping from named slots to spans of text within a document. Documents may feature zero or more instances of a template of any given type, and the task of template extraction entails identifying the templates in a document and extracting each template{‘}s slot values. Our imitation learning approach casts the problem as a Markov decision process (MDP), and relieves the need to use predefined template orders to train an extractor. It leads to state-of-the-art results on two established benchmarks {–} 4-ary relation extraction on SciREX and template extraction on MUC-4 {–} as well as a strong baseline on the new BETTER Granular task.
@inproceedings{chen-etal-2023-iterative, title = "Iterative Document-level Information Extraction via Imitation Learning", author = "Chen, Yunmo and Gantt, William and Gu, Weiwei and Chen, Tongfei and White, Aaron and Van Durme, Benjamin", editor = "Vlachos, Andreas and Augenstein, Isabelle", booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics", month = may, year = "2023", address = "Dubrovnik, Croatia", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.eacl-main.136", doi = "10.18653/v1/2023.eacl-main.136", pages = "1858--1874", abstract = "We present a novel iterative extraction model, IterX, for extracting complex relations, or templates, i.e., N-tuples representing a mapping from named slots to spans of text within a document. Documents may feature zero or more instances of a template of any given type, and the task of template extraction entails identifying the templates in a document and extracting each template{'}s slot values. Our imitation learning approach casts the problem as a Markov decision process (MDP), and relieves the need to use predefined template orders to train an extractor. It leads to state-of-the-art results on two established benchmarks {--} 4-ary relation extraction on SciREX and template extraction on MUC-4 {--} as well as a strong baseline on the new BETTER Granular task.", } - P. Abadir, Ramalingam Chellappa, N. Choudhry, G. Demiris, Deepak Ganesan, Jason Karlawish, Rose M Li, Jason H. Moore, J. Walston, Benjamin Najim Alicia I. Mathias Thomas K. M. Suchi Esther Marlin Dehak Arbaje Unberath Cudjoe Saria Oh Lunde, Benjamin M Marlin, N. Dehak, A. Arbaje, M. Unberath, T. Cudjoe, S. Saria, Esther Oh, N. Lundebjerg, C. Chute, Phillip Phan, Quincy M. Samus, and Nancy L. Schoenborn, “The promise of AI and technology to improve quality of life and care for older adults,” in Nature Aging, 2023.
[BibTeX] [Link]@inproceedings{258909084, title = {The promise of AI and technology to improve quality of life and care for older adults}, author = {{P. Abadir} and {Ramalingam Chellappa} and {N. Choudhry} and {G. Demiris} and {Deepak Ganesan} and {Jason Karlawish} and {Rose M Li} and {Jason H. Moore} and {J. Walston} and {Benjamin Najim Alicia I. Mathias Thomas K. M. Suchi Esther Marlin Dehak Arbaje Unberath Cudjoe Saria Oh Lunde} and {Benjamin M Marlin} and {N. Dehak} and {A. Arbaje} and {M. Unberath} and {T. Cudjoe} and {S. Saria} and {Esther Oh} and {N. Lundebjerg} and {C. Chute} and {Phillip Phan} and {Quincy M. Samus} and {Nancy L. Schoenborn}}, year = 2023, month = {5}, booktitle = {Nature Aging}, url = {https://www.semanticscholar.org/paper/24eafaf005bd6d73870b66525e8978b760e7b3ad}, } - Qihao Liu, Adam Kortylewski, and A. Yuille, “PoseExaminer: Automated Testing of Out-of-Distribution Robustness in Human Pose and Shape Estimation,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{257495961, title = {PoseExaminer: Automated Testing of Out-of-Distribution Robustness in Human Pose and Shape Estimation}, author = {{Qihao Liu} and {Adam Kortylewski} and {A. Yuille}}, year = 2023, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/85fcce7ef6f5eec2d5e5bce82fc7246e8a90696c}, } - A. Favaro, L. Moro-Velázquez, A. Butala, C. Motley, Tianyu Cao, R. Stevens, J. Villalba, and N. Dehak, “Multilingual evaluation of interpretable biomarkers to represent language and speech patterns in Parkinson’s disease,” in Frontiers in Neurology, 2023.
[BibTeX] [Link]@inproceedings{257323163, title = {Multilingual evaluation of interpretable biomarkers to represent language and speech patterns in Parkinson's disease}, author = {{A. Favaro} and {L. Moro-Velázquez} and {A. Butala} and {C. Motley} and {Tianyu Cao} and {R. Stevens} and {J. Villalba} and {N. Dehak}}, year = 2023, month = {3}, booktitle = {Frontiers in Neurology}, url = {https://www.semanticscholar.org/paper/3ed2d557a323c9fc39dbdd64e0ffab064b35a7f9}, } - Rachael E. Jack, Vishal M. Patel, P. Turaga, Mayank Vatsa, Ramalingam Chellappa, A. Pentland, and Richa Singh, “Best Paper Section IEEE International Conference on Automatic Face and Gesture Recognition 2021,” in IEEE Transactions on Biometrics Behavior and Identity Science, 2023.
[BibTeX] [Link]@inproceedings{260715129, title = {Best Paper Section IEEE International Conference on Automatic Face and Gesture Recognition 2021}, author = {{Rachael E. Jack} and {Vishal M. Patel} and {P. Turaga} and {Mayank Vatsa} and {Ramalingam Chellappa} and {A. Pentland} and {Richa Singh}}, year = 2023, month = {7}, booktitle = {IEEE Transactions on Biometrics Behavior and Identity Science}, url = {https://www.semanticscholar.org/paper/51bdfb0e1834a150abd52ad73b63e80ad690aa80}, } - Yining Lu, Haoping Yu, and Daniel Khashabi, “GEAR: Augmenting Language Models with Generalizable and Efficient Tool Resolution,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259950912, title = {GEAR: Augmenting Language Models with Generalizable and Efficient Tool Resolution}, author = {{Yining Lu} and {Haoping Yu} and {Daniel Khashabi}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3bd83ff979f3c0e9470f23c360a18333593dc5a1}, } - Orion Weller, Kyle Lo, David Wadden, Dawn J Lawrie, Benjamin Van Durme, Arman Cohan, and Luca Soldaini, “When do Generative Query and Document Expansions Fail? A Comprehensive Study Across Methods, Retrievers, and Datasets,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{262012661, title = {When do Generative Query and Document Expansions Fail? A Comprehensive Study Across Methods, Retrievers, and Datasets}, author = {{Orion Weller} and {Kyle Lo} and {David Wadden} and {Dawn J Lawrie} and {Benjamin Van Durme} and {Arman Cohan} and {Luca Soldaini}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/dcba37400ad86aaf0d47e5cdfa2fcfa98f089401}, } - Matt Post and Marcin Junczys-Dowmunt, “Escaping the sentence-level paradigm in machine translation,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258309151, title = {Escaping the sentence-level paradigm in machine translation}, author = {{Matt Post} and {Marcin Junczys-Dowmunt}}, year = 2023, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/59e2cfbb1395a4a02e9efecadd4f4005af462c1b}, } - Chen Wei, K. Mangalam, Po-Yao (Bernie) Huang, Yanghao Li, Haoqi Fan, Hu Xu, Huiyu Wang, Cihang Xie, A. Yuille, and Christoph Feichtenhofer, “Diffusion Models as Masked Autoencoders,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257985028, title = {Diffusion Models as Masked Autoencoders}, author = {{Chen Wei} and {K. Mangalam} and {Po-Yao (Bernie) Huang} and {Yanghao Li} and {Haoqi Fan} and {Hu Xu} and {Huiyu Wang} and {Cihang Xie} and {A. Yuille} and {Christoph Feichtenhofer}}, year = 2023, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b032f324a0d4a24fd917551345bd100dc368e41a}, } - Nisarg A. Shah, S. Sikder, S. Vedula, and Vishal M. Patel, “GLSFormer: Gated – Long, Short Sequence Transformer for Step Recognition in Surgical Videos,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259991385, title = {GLSFormer: Gated - Long, Short Sequence Transformer for Step Recognition in Surgical Videos}, author = {{Nisarg A. Shah} and {S. Sikder} and {S. Vedula} and {Vishal M. Patel}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/330bf5c3162606581ebfba1f744e1f7da90c5de4}, } - Artur Jesslen, Guofeng Zhang, Angtian Wang, A. Yuille, and Adam Kortylewski, “Robust 3D-aware Object Classification via Discriminative Render-and-Compare,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258865176, title = {Robust 3D-aware Object Classification via Discriminative Render-and-Compare}, author = {{Artur Jesslen} and {Guofeng Zhang} and {Angtian Wang} and {A. Yuille} and {Adam Kortylewski}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/add9ac3eab96f301f2be5cd57c7c27eb38664376}, } - Elias Stengel-Eskin and Benjamin Van Durme, “Did You Mean…? Confidence-based Trade-offs in Semantic Parsing,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257805227, title = {Did You Mean...? Confidence-based Trade-offs in Semantic Parsing}, author = {{Elias Stengel-Eskin} and {Benjamin Van Durme}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/695a24f4bd79293d7c4dc41ce3f86c66d601f930}, } - V. Raunak, A. Menezes, M. Post, and H. Hassan, “Do GPTs Produce Less Literal Translations?,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Toronto, Canada, 2023, p. 1041–1050. doi:10.18653/v1/2023.acl-short.90
[BibTeX] [Abstract] [Link]
Large Language Models (LLMs) such as GPT-3 have emerged as general-purpose language models capable of addressing many natural language generation or understanding tasks. On the task of Machine Translation (MT), multiple works have investigated few-shot prompting mechanisms to elicit better translations from LLMs. However, there has been relatively little investigation on how such translations differ qualitatively from the translations generated by standard Neural Machine Translation (NMT) models. In this work, we investigate these differences in terms of the literalness of translations produced by the two systems. Using literalness measures involving word alignment and monotonicity, we find that translations out of English (E-X) from GPTs tend to be less literal, while exhibiting similar or better scores on MT quality metrics. We demonstrate that this finding is borne out in human evaluations as well. We then show that these differences are especially pronounced when translating sentences that contain idiomatic expressions.
@inproceedings{raunak-etal-2023-gpts, title = "Do {GPT}s Produce Less Literal Translations?", author = "Raunak, Vikas and Menezes, Arul and Post, Matt and Hassan, Hany", editor = "Rogers, Anna and Boyd-Graber, Jordan and Okazaki, Naoaki", booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", month = jul, year = "2023", address = "Toronto, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.acl-short.90", doi = "10.18653/v1/2023.acl-short.90", pages = "1041--1050", abstract = "Large Language Models (LLMs) such as GPT-3 have emerged as general-purpose language models capable of addressing many natural language generation or understanding tasks. On the task of Machine Translation (MT), multiple works have investigated few-shot prompting mechanisms to elicit better translations from LLMs. However, there has been relatively little investigation on how such translations differ qualitatively from the translations generated by standard Neural Machine Translation (NMT) models. In this work, we investigate these differences in terms of the literalness of translations produced by the two systems. Using literalness measures involving word alignment and monotonicity, we find that translations out of English (E-X) from GPTs tend to be less literal, while exhibiting similar or better scores on MT quality metrics. We demonstrate that this finding is borne out in human evaluations as well. We then show that these differences are especially pronounced when translating sentences that contain idiomatic expressions.", } - Malsha V. Perera and Vishal M. Patel, “Analyzing Bias in Diffusion-based Face Generation Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258615767, title = {Analyzing Bias in Diffusion-based Face Generation Models}, author = {{Malsha V. Perera} and {Vishal M. Patel}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/94831cbd104369092b08f3711e6ac95c5f5f2c7b}, } - Saurabhchand Bhati, J. Villalba, Piotr Żelasko, L. Moro-Velázquez, and N. Dehak, “Regularizing Contrastive Predictive Coding for Speech Applications.” 2023.
[BibTeX] [Link]@inproceedings{258079344, title = {Regularizing Contrastive Predictive Coding for Speech Applications}, author = {{Saurabhchand Bhati} and {J. Villalba} and {Piotr Żelasko} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2023, month = {4}, booktitle = {}, url = {https://www.semanticscholar.org/paper/47ac48e7ee37e7cf4d3bb183477e42d6c5632b64}, } - Enayat Ullah and R. Arora, “Generalization bounds for Kernel Canonical Correlation Analysis,” in Trans. Mach. Learn. Res., 2023.
[BibTeX] [Link]@inproceedings{258766137, title = {Generalization bounds for Kernel Canonical Correlation Analysis}, author = {{Enayat Ullah} and {R. Arora}}, year = 2023, booktitle = {Trans. Mach. Learn. Res.}, url = {https://www.semanticscholar.org/paper/4a55079d0145870461cbe2a48f53e40e64b7db3d}, } - Chen Wang, Angtian Wang, Junbo Li, A. Yuille, and Cihang Xie, “Benchmarking Robustness in Neural Radiance Fields,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{255569926, title = {Benchmarking Robustness in Neural Radiance Fields}, author = {{Chen Wang} and {Angtian Wang} and {Junbo Li} and {A. Yuille} and {Cihang Xie}}, year = 2023, month = {1}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/eaf0c04e9784d6efc8f9ce16d1d9c3ae43506ad9}, } - Yixiong Chen, A. Yuille, and Zongwei Zhou, “Which Layer is Learning Faster? A Systematic Exploration of Layer-wise Convergence Rate for Deep Neural Networks,” in International Conference on Learning Representations, 2023.
[BibTeX] [Link]@inproceedings{259298670, title = {Which Layer is Learning Faster? A Systematic Exploration of Layer-wise Convergence Rate for Deep Neural Networks}, author = {{Yixiong Chen} and {A. Yuille} and {Zongwei Zhou}}, year = 2023, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/ca195fc6d5879060135a5cf6ff571b243f0c6156}, } - A. Hussein, Dorsa Zeinali, Ondrej Klejch, Matthew Wiesner, Brian Yan, Shammur A. Chowdhury, Ahmed Ali, Shinji Watanabe, and S. Khudanpur, “Speech collage: code-switched audio generation by collaging monolingual corpora,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263152081, title = {Speech collage: code-switched audio generation by collaging monolingual corpora}, author = {{A. Hussein} and {Dorsa Zeinali} and {Ondrej Klejch} and {Matthew Wiesner} and {Brian Yan} and {Shammur A. Chowdhury} and {Ahmed Ali} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/fa5ebb425c57f6c4f1c36a7200ef1da867346e8c}, } - Ashwin Bellur, Karan Thakkar, and Mounya Elhilali, “Explicit-memory multiresolution adaptive framework for speech and music separation,” in EURASIP Journal on Audio, Speech, and Music Processing, 2023.
[BibTeX] [Link]@inproceedings{258570384, title = {Explicit-memory multiresolution adaptive framework for speech and music separation}, author = {{Ashwin Bellur} and {Karan Thakkar} and {Mounya Elhilali}}, year = 2023, month = {5}, booktitle = {EURASIP Journal on Audio, Speech, and Music Processing}, url = {https://www.semanticscholar.org/paper/237ea0d3b14b924f12693a29de6fa903a3ae54ed}, } - S. Sia and K. Duh, “In-context Learning as Maintaining Coherency: A Study of On-the-fly Machine Translation Using Large Language Models,” in Proceedings of Machine Translation Summit XIX, Vol. 1: Research Track, Macau SAR, China, 2023, p. 173–185.
[BibTeX] [Abstract] [Link]
The phenomena of in-context learning has typically been thought of as {“}learning from examples{”}. In this work which focuses on Machine Translation, we present a perspective of in-context learning as the desired generation task maintaining coherency with its context, i.e., the prompt examples. We first investigate randomly sampled prompts across 4 domains, and find that translation performance improves when shown in-domain prompts. Next, we investigate coherency for the in-domain setting, which uses prompt examples from a moving window. We study this with respect to other factors that have previously been identified in the literature such as length, surface similarity and sentence embedding similarity. Our results across 3 models (GPTNeo2.7B, Bloom3B, XGLM2.9B), and three translation directions (en$\rightarrow${pt, de, fr}) suggest that the long-term coherency of the prompts and the test sentence is a good indicator of downstream translation performance. In doing so, we demonstrate the efficacy of in-context Machine Translation for on-the-fly adaptation.
@inproceedings{sia-duh-2023-context, title = "In-context Learning as Maintaining Coherency: A Study of On-the-fly Machine Translation Using Large Language Models", author = "Sia, Suzanna and Duh, Kevin", editor = "Utiyama, Masao and Wang, Rui", booktitle = "Proceedings of Machine Translation Summit XIX, Vol. 1: Research Track", month = sep, year = "2023", address = "Macau SAR, China", publisher = "Asia-Pacific Association for Machine Translation", url = "https://aclanthology.org/2023.mtsummit-research.15", pages = "173--185", abstract = "The phenomena of in-context learning has typically been thought of as {``}learning from examples{''}. In this work which focuses on Machine Translation, we present a perspective of in-context learning as the desired generation task maintaining coherency with its context, i.e., the prompt examples. We first investigate randomly sampled prompts across 4 domains, and find that translation performance improves when shown in-domain prompts. Next, we investigate coherency for the in-domain setting, which uses prompt examples from a moving window. We study this with respect to other factors that have previously been identified in the literature such as length, surface similarity and sentence embedding similarity. Our results across 3 models (GPTNeo2.7B, Bloom3B, XGLM2.9B), and three translation directions (en$\rightarrow${pt, de, fr}) suggest that the long-term coherency of the prompts and the test sentence is a good indicator of downstream translation performance. In doing so, we demonstrate the efficacy of in-context Machine Translation for on-the-fly adaptation.", } - Yuxiang Guo, Cheng-Fang Peng, R. Prabhakar, Chun Pong Lau, and R. Chellappa, “GADER: GAit DEtection and Recognition in the Wild,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{260202959, title = {GADER: GAit DEtection and Recognition in the Wild}, author = {{Yuxiang Guo} and {Cheng-Fang Peng} and {R. Prabhakar} and {Chun Pong Lau} and {R. Chellappa}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/883f84e5fd894cec4b0364999a0461534f048cee}, } - Elias Stengel-Eskin, Kyle Rawlins, and Benjamin Van Durme, “Zero and Few-shot Semantic Parsing with Ambiguous Inputs,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258999294, title = {Zero and Few-shot Semantic Parsing with Ambiguous Inputs}, author = {{Elias Stengel-Eskin} and {Kyle Rawlins} and {Benjamin Van Durme}}, year = 2023, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b8e49068441a43aaf039527c6063a033368dd357}, } - Andrew Blair-Stanek, Nils Holzenberger, and Benjamin Van Durme, “OpenAI Cribbed Our Tax Example, But Can GPT-4 Really Do Tax?,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{262056669, title = {OpenAI Cribbed Our Tax Example, But Can GPT-4 Really Do Tax?}, author = {{Andrew Blair-Stanek} and {Nils Holzenberger} and {Benjamin Van Durme}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b7cf038a3b4c1456a07db0564d940ad8f8b75ada}, } - Margaret A. McMullin, N. Higgins, Brian Gygi, Rohit Kumar, Mounya Elhilali, and J. Snyder, “Perception of global properties, objects, and settings in natural auditory scenes,” in Journal of the Acoustical Society of America, 2023.
[BibTeX] [Link]@inproceedings{258374319, title = {Perception of global properties, objects, and settings in natural auditory scenes}, author = {{Margaret A. McMullin} and {N. Higgins} and {Brian Gygi} and {Rohit Kumar} and {Mounya Elhilali} and {J. Snyder}}, year = 2023, month = {3}, booktitle = {Journal of the Acoustical Society of America}, url = {https://www.semanticscholar.org/paper/8890c3afc837ab0871d11f6ed17bfb6109943bff}, } - N. Higgins, Alexandra N Scurry, Fang Jiang, David F. Little, Claude Alain, Mounya Elhilali, and J. Snyder, “Adaptation in the sensory cortex drives bistable switching during auditory stream segregation,” in Neuroscience of Consciousness, 2023.
[BibTeX] [Link]@inproceedings{256604402, title = {Adaptation in the sensory cortex drives bistable switching during auditory stream segregation}, author = {{N. Higgins} and {Alexandra N Scurry} and {Fang Jiang} and {David F. Little} and {Claude Alain} and {Mounya Elhilali} and {J. Snyder}}, year = 2023, month = {1}, booktitle = {Neuroscience of Consciousness}, url = {https://www.semanticscholar.org/paper/c1c4a48270174de06f609bb2dc98c8e896ce78a3}, } - Dongji Gao, Matthew Wiesner, Hainan Xu, Leibny Paola Garcia, Daniel Povey, and S. Khudanpur, “Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{259063779, title = {Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts}, author = {{Dongji Gao} and {Matthew Wiesner} and {Hainan Xu} and {Leibny Paola Garcia} and {Daniel Povey} and {S. Khudanpur}}, year = 2023, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/b620d46668d62930e41393168434118bb9a2bfcb}, } - Ruizhe Huang, Matthew Wiesner, Leibny Paola García-Perera, Daniel Povey, J. Trmal, and S. Khudanpur, “Building Keyword Search System from End-To-End Asr Systems,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2023.
[BibTeX] [Link]@inproceedings{258535938, title = {Building Keyword Search System from End-To-End Asr Systems}, author = {{Ruizhe Huang} and {Matthew Wiesner} and {Leibny Paola García-Perera} and {Daniel Povey} and {J. Trmal} and {S. Khudanpur}}, year = 2023, month = {6}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/1b610ce986449cbef77d0f6bdd28421fd8495268}, } - Ju He, Jieneng Chen, Ming-Xian Lin, Qihang Yu, and A. Yuille, “Compositor: Bottom-Up Clustering and Compositing for Robust Part and Object Segmentation,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{259145290, title = {Compositor: Bottom-Up Clustering and Compositing for Robust Part and Object Segmentation}, author = {{Ju He} and {Jieneng Chen} and {Ming-Xian Lin} and {Qihang Yu} and {A. Yuille}}, year = 2023, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/c49666de550031cd63514dacc74b5c4a632da6a6}, } - Drew Prinster, S. Saria, and Anqi Liu, “JAWS-X: Addressing Efficiency Bottlenecks of Conformal Prediction Under Standard and Feedback Covariate Shift,” in International Conference on Machine Learning, 2023.
[BibTeX] [Link]@inproceedings{259300311, title = {JAWS-X: Addressing Efficiency Bottlenecks of Conformal Prediction Under Standard and Feedback Covariate Shift}, author = {{Drew Prinster} and {S. Saria} and {Anqi Liu}}, year = 2023, booktitle = {International Conference on Machine Learning}, url = {https://www.semanticscholar.org/paper/6faf60347b9ec3672a4d191cfe9fe0076191e9a0}, } - Kate Sanders, David Etter, Reno Kriz, and Benjamin Van Durme, “MultiVENT: Multilingual Videos of Events with Aligned Natural Text,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259361037, title = {MultiVENT: Multilingual Videos of Events with Aligned Natural Text}, author = {{Kate Sanders} and {David Etter} and {Reno Kriz} and {Benjamin Van Durme}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d6592cd383da0ebae9f4ff78a4df96ab635acfa2}, } - Haoran Xu, Maha Elbayad, Kenton Murray, Jean Maillard, and Vedanuj Goswami, “Towards Being Parameter-Efficient: A Stratified Sparsely Activated Transformer with Dynamic Capacity,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258460982, title = {Towards Being Parameter-Efficient: A Stratified Sparsely Activated Transformer with Dynamic Capacity}, author = {{Haoran Xu} and {Maha Elbayad} and {Kenton Murray} and {Jean Maillard} and {Vedanuj Goswami}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/09d3a354d22d75deddb0dcc99870f307aae2fd3c}, } - Thanh Nguyen-Tang and R. Arora, “Provably Efficient Neural Offline Reinforcement Learning via Perturbed Rewards,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257206027, title = {Provably Efficient Neural Offline Reinforcement Learning via Perturbed Rewards}, author = {{Thanh Nguyen-Tang} and {R. Arora}}, year = 2023, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/a8dac0d0837ac4800f4462a121c59a98a05531ee}, } - Matt Post, Thamme Gowda, Roman Grundkiewicz, Huda Khayrallah, Rohit Jain, and Marcin Junczys-Dowmunt, “SOTASTREAM: A Streaming Approach to Machine Translation Training,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{260900231, title = {SOTASTREAM: A Streaming Approach to Machine Translation Training}, author = {{Matt Post} and {Thamme Gowda} and {Roman Grundkiewicz} and {Huda Khayrallah} and {Rohit Jain} and {Marcin Junczys-Dowmunt}}, year = 2023, month = {8}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/ea6792cd3acbe38954ead38f222448457db19347}, } - Jeya Maria Jose Valanarasu, Yucheng Tang, Dong Yang, Ziyue Xu, Can Zhao, Wenqi Li, Vishal M. Patel, Bennett A. Landman, Daguang Xu, Yufan He, and V. Nath, “Disruptive Autoencoders: Leveraging Low-level features for 3D Medical Image Pre-training,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{260334568, title = {Disruptive Autoencoders: Leveraging Low-level features for 3D Medical Image Pre-training}, author = {{Jeya Maria Jose Valanarasu} and {Yucheng Tang} and {Dong Yang} and {Ziyue Xu} and {Can Zhao} and {Wenqi Li} and {Vishal M. Patel} and {Bennett A. Landman} and {Daguang Xu} and {Yufan He} and {V. Nath}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/21510551199cc8ebfba5252575133de607448fa9}, } - Orion Weller, Dawn J Lawrie, and Benjamin Van Durme, “NevIR: Negation in Neural Information Retrieval,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258676146, title = {NevIR: Negation in Neural Information Retrieval}, author = {{Orion Weller} and {Dawn J Lawrie} and {Benjamin Van Durme}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/402cbd67c202b0b2dc184861472bd99ebc0e69a4}, } - R. Chellappa, “The unsung hero: how synthetic data has helped computer vision, machine learning, and AI,” in Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications, 2023.
[BibTeX] [Link]@inproceedings{259436318, title = {The unsung hero: how synthetic data has helped computer vision, machine learning, and AI}, author = {{R. Chellappa}}, year = 2023, month = {6}, booktitle = {Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications}, url = {https://www.semanticscholar.org/paper/c061dd875146aa8d87b5bfe45eea73df8da3c373}, } - Thanh Nguyen-Tang and R. Arora, “VIPeR: Provably Efficient Algorithm for Offline RL with Neural Function Approximation,” in International Conference on Learning Representations, 2023.
[BibTeX] [Link]@inproceedings{257366012, title = {VIPeR: Provably Efficient Algorithm for Offline RL with Neural Function Approximation}, author = {{Thanh Nguyen-Tang} and {R. Arora}}, year = 2023, month = {2}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/7d200b868cb92657a68ac64c112a2cd0a4045f87}, } - Yi Zhang, Pengliang Ji, Angtian Wang, Jieru Mei, Adam Kortylewski, and A. Yuille, “3D-Aware Neural Body Fitting for Occlusion Robust 3D Human Pose Estimation,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{261048725, title = {3D-Aware Neural Body Fitting for Occlusion Robust 3D Human Pose Estimation}, author = {{Yi Zhang} and {Pengliang Ji} and {Angtian Wang} and {Jieru Mei} and {Adam Kortylewski} and {A. Yuille}}, year = 2023, month = {8}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/8a46058e5353ae5cc66dd8fc67e635aaad3a5f17}, } - W. G. C. Bandara and Vishal M. Patel, “Deep Metric Learning for Unsupervised Remote Sensing Change Detection,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257557771, title = {Deep Metric Learning for Unsupervised Remote Sensing Change Detection}, author = {{W. G. C. Bandara} and {Vishal M. Patel}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/30d02457a38374398deca536682c193f0f0b1a24}, } - Andrew Blair-Stanek, Nils Holzenberger, and Benjamin Van Durme, “Can GPT-3 Perform Statutory Reasoning?,” in International Conference on Artificial Intelligence and Law, 2023.
[BibTeX] [Link]@inproceedings{256826996, title = {Can GPT-3 Perform Statutory Reasoning?}, author = {{Andrew Blair-Stanek} and {Nils Holzenberger} and {Benjamin Van Durme}}, year = 2023, month = {2}, booktitle = {International Conference on Artificial Intelligence and Law}, url = {https://www.semanticscholar.org/paper/5f5253fb15ac382e96ade0335baf1cfaa240fb1d}, } - VS Vibashan, Ning Yu, Chen Xing, Can Qin, M. Gao, Juan Carlos Niebles, Vishal M. Patel, and Ran Xu, “Mask-Free OVIS: Open-Vocabulary Instance Segmentation without Manual Mask Annotations,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{257804958, title = {Mask-Free OVIS: Open-Vocabulary Instance Segmentation without Manual Mask Annotations}, author = {{VS Vibashan} and {Ning Yu} and {Chen Xing} and {Can Qin} and {M. Gao} and {Juan Carlos Niebles} and {Vishal M. Patel} and {Ran Xu}}, year = 2023, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/7aa528b8732033cfb7a6d130bb321723a4e49700}, } - Saurabhchand Bhati, J. Villalba, L. Moro-Velázquez, Thomas Thebaud, and N. Dehak, “Segmental SpeechCLIP: Utilizing Pretrained Image-text Models for Audio-Visual Learning,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{260909100, title = {Segmental SpeechCLIP: Utilizing Pretrained Image-text Models for Audio-Visual Learning}, author = {{Saurabhchand Bhati} and {J. Villalba} and {L. Moro-Velázquez} and {Thomas Thebaud} and {N. Dehak}}, year = 2023, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/1617d389b7947161f2943e2d30afeb1856052b14}, } - A. Favaro, Yi-Ting Tsai, A. Butala, Thomas Thebaud, J. Villalba, N. Dehak, and L. Moro-Velázquez, “Interpretable Speech Features vs. DNN Embeddings: What to Use in the Automatic Assessment of Parkinson’s Disease in Multi-lingual Scenarios,” in medRxiv, 2023.
[BibTeX] [Link]@inproceedings{259047417, title = {Interpretable Speech Features vs. DNN Embeddings: What to Use in the Automatic Assessment of Parkinson's Disease in Multi-lingual Scenarios}, author = {{A. Favaro} and {Yi-Ting Tsai} and {A. Butala} and {Thomas Thebaud} and {J. Villalba} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2023, month = {6}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/8d18efe22ad66b53a0a13fc71c9b57c41b7790d0}, } - S. Styles, Victoria Y. H. Chua, Fei Ting Woon, Hexin Liu, Leibny Paola Garcia Perera, S. Khudanpur, Andy W. H. Khong, and J. Dauwels, “Investigating model performance in language identification: beyond simple error statistics,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{258967563, title = {Investigating model performance in language identification: beyond simple error statistics}, author = {{S. Styles} and {Victoria Y. H. Chua} and {Fei Ting Woon} and {Hexin Liu} and {Leibny Paola Garcia Perera} and {S. Khudanpur} and {Andy W. H. Khong} and {J. Dauwels}}, year = 2023, month = {5}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/5550d0db7030aa77480bfb10c4ec00862bb233eb}, } - Dongji Gao, Hainan Xu, Desh Raj, Leibny Paola Garcia Perera, Daniel Povey, and S. Khudanpur, “Learning from Flawed Data: Weakly Supervised Automatic Speech Recognition,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263152210, title = {Learning from Flawed Data: Weakly Supervised Automatic Speech Recognition}, author = {{Dongji Gao} and {Hainan Xu} and {Desh Raj} and {Leibny Paola Garcia Perera} and {Daniel Povey} and {S. Khudanpur}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/4df2d56e2c81d315e8ead7c3eaf840064ea3665e}, } - J. Villalba, Jonas Borgstrom, Maliha Jahan, Saurabh Kataria, Leibny Paola Garcia, P. Torres-Carrasquillo, and N. Dehak, “Advances in Language Recognition in Low Resource African Languages: The JHU-MIT Submission for NIST LRE22,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{260918551, title = {Advances in Language Recognition in Low Resource African Languages: The JHU-MIT Submission for NIST LRE22}, author = {{J. Villalba} and {Jonas Borgstrom} and {Maliha Jahan} and {Saurabh Kataria} and {Leibny Paola Garcia} and {P. Torres-Carrasquillo} and {N. Dehak}}, year = 2023, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/51aca07d500c44ebde896b8df3b0388dd3ade489}, } - L. Eudy and Matt Post, “AC Transit Fuel Cell Electric Bus Progress Report (Data Period Focus: Jan. 2020 through Dec. 2020).” 2023.
[BibTeX] [Link]@inproceedings{257848515, title = {AC Transit Fuel Cell Electric Bus Progress Report (Data Period Focus: Jan. 2020 through Dec. 2020)}, author = {{L. Eudy} and {Matt Post}}, year = 2023, month = {3}, booktitle = {}, url = {https://www.semanticscholar.org/paper/69cf8aae9aa20f261b91ad67636fc064a2376e7a}, } - Victoria Y. H. Chua, Hexin Liu, Leibny Paola Garcia Perera, Fei Ting Woon, Jinyi Wong, Xiangyu Zhang, S. Khudanpur, Andy W. H. Khong, J. Dauwels, and S. Styles, “MERLIon CCS Challenge: A English-Mandarin code-switching child-directed speech corpus for language identification and diarization,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{258967838, title = {MERLIon CCS Challenge: A English-Mandarin code-switching child-directed speech corpus for language identification and diarization}, author = {{Victoria Y. H. Chua} and {Hexin Liu} and {Leibny Paola Garcia Perera} and {Fei Ting Woon} and {Jinyi Wong} and {Xiangyu Zhang} and {S. Khudanpur} and {Andy W. H. Khong} and {J. Dauwels} and {S. Styles}}, year = 2023, month = {5}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/ea701359c6e2b2ba5c36c4849d4144d318171418}, } - Sai Saketh Rambhatla, Ishan Misra, R. Chellappa, and Abhinav Shrivastava, “MOST: Multiple Object localization with Self-supervised Transformers for object discovery,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258060050, title = {MOST: Multiple Object localization with Self-supervised Transformers for object discovery}, author = {{Sai Saketh Rambhatla} and {Ishan Misra} and {R. Chellappa} and {Abhinav Shrivastava}}, year = 2023, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3be837073f08eecc01e1bc742c541c5f0e644946}, } - Yu Zeng, Mo Zhou, Yuan Xue, and Vishal M. Patel, “Securing Deep Generative Models with Universal Adversarial Signature,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258888240, title = {Securing Deep Generative Models with Universal Adversarial Signature}, author = {{Yu Zeng} and {Mo Zhou} and {Yuan Xue} and {Vishal M. Patel}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/98e87cd4c19dad7018270b4561dc64b0109ee360}, } - Tianjian Li and Kenton Murray, “Why Does Zero-Shot Cross-Lingual Generation Fail? An Explanation and a Solution,” in Annual Meeting of the Association for Computational Linguistics, 2023.
[BibTeX] [Link]@inproceedings{258959429, title = {Why Does Zero-Shot Cross-Lingual Generation Fail? An Explanation and a Solution}, author = {{Tianjian Li} and {Kenton Murray}}, year = 2023, month = {5}, booktitle = {Annual Meeting of the Association for Computational Linguistics}, url = {https://www.semanticscholar.org/paper/aa933e27c470eeecbe7bbec5debdd8c5d2faa4be}, } - Anjalie Field, Prateek Verma, Nay San, J. Eberhardt, and Dan Jurafsky, “Developing Speech Processing Pipelines for Police Accountability,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{259129355, title = {Developing Speech Processing Pipelines for Police Accountability}, author = {{Anjalie Field} and {Prateek Verma} and {Nay San} and {J. Eberhardt} and {Dan Jurafsky}}, year = 2023, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/38c2e4e54a50ea5027c7a06ab325c22ece7b6c40}, } - Desh Raj, Daniel Povey, and S. Khudanpur, “SURT 2.0: Advances in Transducer-Based Multi-Talker Speech Recognition,” in IEEE/ACM Transactions on Audio Speech and Language Processing, 2023.
[BibTeX] [Link]@inproceedings{259202526, title = {SURT 2.0: Advances in Transducer-Based Multi-Talker Speech Recognition}, author = {{Desh Raj} and {Daniel Povey} and {S. Khudanpur}}, year = 2023, month = {6}, booktitle = {IEEE/ACM Transactions on Audio Speech and Language Processing}, url = {https://www.semanticscholar.org/paper/a6fffd418fabef307cba5e70324a3ba89c7ffc39}, } - Jay N. Paranjape, S. Sikder, Vishal M. Patel, and S. Vedula, “Cross-Dataset Adaptation for Instrument Classification in Cataract Surgery Videos,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2023.
[BibTeX] [Link]@inproceedings{260704213, title = {Cross-Dataset Adaptation for Instrument Classification in Cataract Surgery Videos}, author = {{Jay N. Paranjape} and {S. Sikder} and {Vishal M. Patel} and {S. Vedula}}, year = 2023, month = {7}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/648a8bc5f5e6354aa56a899c327715bfaa80944d}, } - Jonah P. Sengupta and A. Andreou, “Retinomorphic Channel Design and Considerations,” in Annual Conference on Information Sciences and Systems, 2023.
[BibTeX] [Link]@inproceedings{258074434, title = {Retinomorphic Channel Design and Considerations}, author = {{Jonah P. Sengupta} and {A. Andreou}}, year = 2023, month = {3}, booktitle = {Annual Conference on Information Sciences and Systems}, url = {https://www.semanticscholar.org/paper/7f97effeed913a6089ca98d576d585401e251f9b}, } - Chongyu Qu, Tiezheng Zhang, Hualin Qiao, Jie Liu, Yucheng Tang, A. Yuille, and Zongwei Zhou, “AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weeks.” 2023.
[BibTeX] [Link]@inproceedings{258714648, title = {AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weeks}, author = {{Chongyu Qu} and {Tiezheng Zhang} and {Hualin Qiao} and {Jie Liu} and {Yucheng Tang} and {A. Yuille} and {Zongwei Zhou}}, year = 2023, month = {5}, booktitle = {}, url = {https://www.semanticscholar.org/paper/40927a6916f00f8e7ad43ad7009decb25a861767}, } - Y. Ranasinghe, Nithin Gopalakrishnan Nair, W. G. C. Bandara, and Vishal M. Patel, “Diffuse-Denoise-Count: Accurate Crowd-Counting with Diffusion Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257663507, title = {Diffuse-Denoise-Count: Accurate Crowd-Counting with Diffusion Models}, author = {{Y. Ranasinghe} and {Nithin Gopalakrishnan Nair} and {W. G. C. Bandara} and {Vishal M. Patel}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/851a11cfb1a5832f81717ee9a5b03901fe4f39e9}, } - K. Duh and X. Zhang, “AutoML for NLP,” in Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts, Dubrovnik, Croatia, 2023, p. 25–26. doi:10.18653/v1/2023.eacl-tutorials.5
[BibTeX] [Abstract] [Link]
Automated Machine Learning (AutoML) is an emerging field that has potential to impact how we build models in NLP. As an umbrella term that includes topics like hyperparameter optimization and neural architecture search, AutoML has recently become mainstream at major conferences such as NeurIPS, ICML, and ICLR. What does this mean to NLP? Currently, models are often built in an ad hoc process: we might borrow default hyperparameters from previous work and try a few variant architectures, but it is never guaranteed that final trained model is optimal. Automation can introduce rigor in this model-building process. This tutorial will summarize the main AutoML techniques and illustrate how to apply them to improve the NLP model-building process.
@inproceedings{duh-zhang-2023-automl, title = "{A}uto{ML} for {NLP}", author = "Duh, Kevin and Zhang, Xuan", editor = "Zanzotto, Fabio Massimo and Pradhan, Sameer", booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts", month = may, year = "2023", address = "Dubrovnik, Croatia", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.eacl-tutorials.5", doi = "10.18653/v1/2023.eacl-tutorials.5", pages = "25--26", abstract = "Automated Machine Learning (AutoML) is an emerging field that has potential to impact how we build models in NLP. As an umbrella term that includes topics like hyperparameter optimization and neural architecture search, AutoML has recently become mainstream at major conferences such as NeurIPS, ICML, and ICLR. What does this mean to NLP? Currently, models are often built in an ad hoc process: we might borrow default hyperparameters from previous work and try a few variant architectures, but it is never guaranteed that final trained model is optimal. Automation can introduce rigor in this model-building process. This tutorial will summarize the main AutoML techniques and illustrate how to apply them to improve the NLP model-building process.", } - T. Meyer, A. Favaro, Tianyu Cao, A. Butala, E. Oh, C. Motley, P. Irazoqui, N. Dehak, and L. Moro-Velázquez, “Deep Stroop: Using eye tracking and speech processing to characterize people with neurodegenerative disorders while performing the Stroop Test,” in medRxiv, 2023.
[BibTeX] [Link]@inproceedings{258997982, title = {Deep Stroop: Using eye tracking and speech processing to characterize people with neurodegenerative disorders while performing the Stroop Test}, author = {{T. Meyer} and {A. Favaro} and {Tianyu Cao} and {A. Butala} and {E. Oh} and {C. Motley} and {P. Irazoqui} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2023, month = {6}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/172e04d89d89109626cba6a5b2d4d8a736bd145d}, } - Orion Weller, Marc Marone, Nathaniel Weir, Dawn J Lawrie, Daniel Khashabi, and Benjamin Van Durme, “”According to …” Prompting Language Models Improves Quoting from Pre-Training Data,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258832937, title = {"According to ..." Prompting Language Models Improves Quoting from Pre-Training Data}, author = {{Orion Weller} and {Marc Marone} and {Nathaniel Weir} and {Dawn J Lawrie} and {Daniel Khashabi} and {Benjamin Van Durme}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3cf26008c7d425b8e9c33dec7fd633ec8c87bef6}, } - M. Marín-Jiménez, Shiqi Yu, Yasushi Makihara, Vishal M. Patel, Maneet Singh, and M. De Marsico, “Editorial Introduction to the Special Issue on Biometrics at a Distance in the Deep Learning Era,” in IEEE Journal on Selected Topics in Signal Processing, 2023.
[BibTeX] [Link]@inproceedings{259893511, title = {Editorial Introduction to the Special Issue on Biometrics at a Distance in the Deep Learning Era}, author = {{M. Marín-Jiménez} and {Shiqi Yu} and {Yasushi Makihara} and {Vishal M. Patel} and {Maneet Singh} and {M. De Marsico}}, year = 2023, month = {5}, booktitle = {IEEE Journal on Selected Topics in Signal Processing}, url = {https://www.semanticscholar.org/paper/df958104e921170b592e27798e18de9b9c892cbd}, } - Helin Wang, Thomas Thebaud, J. Villalba, Myra Sydnor, Becky Lammers, N. Dehak, and L. Moro-Velázquez, “DuTa-VC: A Duration-aware Typical-to-atypical Voice Conversion Approach with Diffusion Probabilistic Model,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{259202766, title = {DuTa-VC: A Duration-aware Typical-to-atypical Voice Conversion Approach with Diffusion Probabilistic Model}, author = {{Helin Wang} and {Thomas Thebaud} and {J. Villalba} and {Myra Sydnor} and {Becky Lammers} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2023, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/bf6339e920466f2dc7dc0da5edde5b3187cf9d0d}, } - Enayat Ullah, Harry Lang, R. Arora, and V. Braverman, “Clustering using Approximate Nearest Neighbour Oracles,” in Trans. Mach. Learn. Res., 2023.
[BibTeX] [Link]@inproceedings{258766141, title = {Clustering using Approximate Nearest Neighbour Oracles}, author = {{Enayat Ullah} and {Harry Lang} and {R. Arora} and {V. Braverman}}, year = 2023, booktitle = {Trans. Mach. Learn. Res.}, url = {https://www.semanticscholar.org/paper/2e864475d80f551d97232f9a6cba079dd128c54d}, } - V. Rennoll, I. McLane, Adebayo A. Eisape, D. Grant, Helena Hahn, Mounya Elhilali, and James E. West, “Electrostatic Acoustic Sensor with an Impedance-Matched Diaphragm Characterized for Body Sound Monitoring.,” in ACS Applied Bio Materials, 2023.
[BibTeX] [Link]@inproceedings{259993644, title = {Electrostatic Acoustic Sensor with an Impedance-Matched Diaphragm Characterized for Body Sound Monitoring.}, author = {{V. Rennoll} and {I. McLane} and {Adebayo A. Eisape} and {D. Grant} and {Helena Hahn} and {Mounya Elhilali} and {James E. West}}, year = 2023, month = {7}, booktitle = {ACS Applied Bio Materials}, url = {https://www.semanticscholar.org/paper/bf5172b246adb601b731618108ba8ce5d1367177}, } - Hexin Liu, Leibny Paola Garcia, Xiangyu Zhang, Andy W. H. Khong, and S. Khudanpur, “Enhancing Code-switching Speech Recognition with Interactive Language Biases,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263310495, title = {Enhancing Code-switching Speech Recognition with Interactive Language Biases}, author = {{Hexin Liu} and {Leibny Paola Garcia} and {Xiangyu Zhang} and {Andy W. H. Khong} and {S. Khudanpur}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/47c14df80f649488c64f5659fa49ad356ff59470}, } - Jonah P. Sengupta, M. A. Tomlinson, Daniel R. Mendat, M. Villemur, and A. Andreou, “Asynchronous, Spatiotemporal Filtering using an Analog Cellular Neural Network Processor,” in International Symposium on Circuits and Systems, 2023.
[BibTeX] [Link]@inproceedings{260003954, title = {Asynchronous, Spatiotemporal Filtering using an Analog Cellular Neural Network Processor}, author = {{Jonah P. Sengupta} and {M. A. Tomlinson} and {Daniel R. Mendat} and {M. Villemur} and {A. Andreou}}, year = 2023, month = {5}, booktitle = {International Symposium on Circuits and Systems}, url = {https://www.semanticscholar.org/paper/30446a1b3ca0fc61c3b672d5a284e0dcb761fe6d}, } - Lingfeng Shen, Weiting Tan, Boyuan Zheng, and Daniel Khashabi, “Flatness-Aware Prompt Selection Improves Accuracy and Sample Efficiency,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258762744, title = {Flatness-Aware Prompt Selection Improves Accuracy and Sample Efficiency}, author = {{Lingfeng Shen} and {Weiting Tan} and {Boyuan Zheng} and {Daniel Khashabi}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b8ba16a107621f760e7830ddaab8c3d5c5ff06b0}, } - A. Favaro, Tianyu Cao, Thomas Thebaud, J. Villalba, A. Butala, N. Dehak, and L. Moro-Velázquez, “Do Phonatory Features Display Robustness to Characterize Parkinsonian Speech Across Corpora?,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{260914548, title = {Do Phonatory Features Display Robustness to Characterize Parkinsonian Speech Across Corpora?}, author = {{A. Favaro} and {Tianyu Cao} and {Thomas Thebaud} and {J. Villalba} and {A. Butala} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2023, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/562d06b0cddb553a76e6b68f6f2ba470a17bb5d4}, } - Liangyu Chen, A. Yuille, and Zongwei Zhou, “Making Your First Choice: To Address Cold Start Problem in Medical Active Learning.” 2023.
[BibTeX] [Link]@inproceedings{259373518, title = {Making Your First Choice: To Address Cold Start Problem in Medical Active Learning}, author = {{Liangyu Chen} and {A. Yuille} and {Zongwei Zhou}}, year = 2023, booktitle = {}, url = {https://www.semanticscholar.org/paper/251516c1549fc4566b801788c932ef1f18f343b3}, } - Jie Liu, Yixiao Zhang, Jieneng Chen, Junfei Xiao, Yongyi Lu, Bennett A. Landman, Yixuan Yuan, A. Yuille, Yucheng Tang, and Zongwei Zhou, “CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{255372928, title = {CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection}, author = {{Jie Liu} and {Yixiao Zhang} and {Jieneng Chen} and {Junfei Xiao} and {Yongyi Lu} and {Bennett A. Landman} and {Yixuan Yuan} and {A. Yuille} and {Yucheng Tang} and {Zongwei Zhou}}, year = 2023, month = {1}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/125632627bfad80c2c688bcbed7f3ee915de7359}, } - Yoav Wald and S. Saria, “Birds of an odd feather: guaranteed out-of-distribution (OOD) novel category detection,” in Conference on Uncertainty in Artificial Intelligence, 2023.
[BibTeX] [Link]@inproceedings{260870320, title = {Birds of an odd feather: guaranteed out-of-distribution (OOD) novel category detection}, author = {{Yoav Wald} and {S. Saria}}, year = 2023, booktitle = {Conference on Uncertainty in Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/90869ce11a670ca4fb568a56a38f3ffe37161121}, } - Yiqing Wang, Zihan Li, Jieru Mei, Zi-Ying Wei, Li Liu, Chen Wang, Shengtian Sang, A. Yuille, Cihang Xie, and Yuyin Zhou, “SwinMM: Masked Multi-view with Swin Transformers for 3D Medical Image Segmentation,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2023.
[BibTeX] [Link]@inproceedings{260125020, title = {SwinMM: Masked Multi-view with Swin Transformers for 3D Medical Image Segmentation}, author = {{Yiqing Wang} and {Zihan Li} and {Jieru Mei} and {Zi-Ying Wei} and {Li Liu} and {Chen Wang} and {Shengtian Sang} and {A. Yuille} and {Cihang Xie} and {Yuyin Zhou}}, year = 2023, month = {7}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/c84e03c8ba7feaa2679a90b1637f3b079be15aa9}, } - Saurabhchand Bhati, J. Villalba, L. Moro-Velázquez, Thomas Thebaud, and N. Dehak, “Leveraging Pretrained Image-text Models for Improving Audio-Visual Learning,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{261682358, title = {Leveraging Pretrained Image-text Models for Improving Audio-Visual Learning}, author = {{Saurabhchand Bhati} and {J. Villalba} and {L. Moro-Velázquez} and {Thomas Thebaud} and {N. Dehak}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/ada7b279876196a283a8379729212338386c7eba}, } - Bowen Li, Yu-Cheng Chou, Shuwen Sun, Hualin Qiao, A. Yuille, and Zongwei Zhou, “Early Detection and Localization of Pancreatic Cancer by Label-Free Tumor Synthesis,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{260681299, title = {Early Detection and Localization of Pancreatic Cancer by Label-Free Tumor Synthesis}, author = {{Bowen Li} and {Yu-Cheng Chou} and {Shuwen Sun} and {Hualin Qiao} and {A. Yuille} and {Zongwei Zhou}}, year = 2023, month = {8}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d556aaf4ac55b6cebb9b889ec7b89b086cb41bff}, } - Ishani Mondal, Michelle Yuan, N. Anandhavelu, Aparna Garimella, Francis Ferraro, Andrew Blair-Stanek, Benjamin Van Durme, and Jordan L. Boyd-Graber, “InteractiveIE: Towards Assessing the Strength of Human-AI Collaboration in Improving the Performance of Information Extraction,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258865888, title = {InteractiveIE: Towards Assessing the Strength of Human-AI Collaboration in Improving the Performance of Information Extraction}, author = {{Ishani Mondal} and {Michelle Yuan} and {N. Anandhavelu} and {Aparna Garimella} and {Francis Ferraro} and {Andrew Blair-Stanek} and {Benjamin Van Durme} and {Jordan L. Boyd-Graber}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/21e0a1324522b39e5cec94885501e906942c43d0}, } - Pengfei Yu, Heng Ji, Shih-Fu Chang, and Kevin Duh, “MULTIMEDIA CURRICULUM LEARNING FOR LANGUAGE ACQUISITION.” 2023.
[BibTeX] [Link]@inproceedings{259923580, title = {MULTIMEDIA CURRICULUM LEARNING FOR LANGUAGE ACQUISITION}, author = {{Pengfei Yu} and {Heng Ji} and {Shih-Fu Chang} and {Kevin Duh}}, year = 2023, booktitle = {}, url = {https://www.semanticscholar.org/paper/7c7d8f106f8cd1bdadfd3b46f6ebb1509cb1be42}, } - Leibny Paola Garcia Perera, Y. H. V. Chua, Hexin Liu, Fei Ting Woon, Andy W. H. Khong, J. Dauwels, S. Khudanpur, and S. Styles, “MERLIon CCS Challenge Evaluation Plan.” 2023.
[BibTeX] [Link]@inproceedings{258987816, title = {MERLIon CCS Challenge Evaluation Plan}, author = {{Leibny Paola Garcia Perera} and {Y. H. V. Chua} and {Hexin Liu} and {Fei Ting Woon} and {Andy W. H. Khong} and {J. Dauwels} and {S. Khudanpur} and {S. Styles}}, year = 2023, month = {5}, booktitle = {}, url = {https://www.semanticscholar.org/paper/6616c330539e1f38b8d80d5aec6eaf0be98f9314}, } - Martin Sustek, Samik Sadhu, L. Burget, H. Hermansky, J. Villalba, L. Moro-Velázquez, and N. Dehak, “Stabilized training of joint energy-based models and their practical applications,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257404851, title = {Stabilized training of joint energy-based models and their practical applications}, author = {{Martin Sustek} and {Samik Sadhu} and {L. Burget} and {H. Hermansky} and {J. Villalba} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/46fd16213979b00e741b926539ad4ba7a1acd1cf}, } - Samuele Cornell, Matthew Wiesner, Shinji Watanabe, Desh Raj, Xuankai Chang, Paola García, Yoshiki Masuyama, Zhongqiu Wang, S. Squartini, and S. Khudanpur, “The CHiME-7 DASR Challenge: Distant Meeting Transcription with Multiple Devices in Diverse Scenarios,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259251557, title = {The CHiME-7 DASR Challenge: Distant Meeting Transcription with Multiple Devices in Diverse Scenarios}, author = {{Samuele Cornell} and {Matthew Wiesner} and {Shinji Watanabe} and {Desh Raj} and {Xuankai Chang} and {Paola García} and {Yoshiki Masuyama} and {Zhongqiu Wang} and {S. Squartini} and {S. Khudanpur}}, year = 2023, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d4d5fe4a35e9de845877015075f727415e83d18f}, } - Anjalie Field, Amanda Coston, Nupoor Gandhi, A. Chouldechova, Emily Putnam-Hornstein, David Steier, and Yulia Tsvetkov, “Examining risks of racial biases in NLP tools for child protective services,” in Conference on Fairness, Accountability and Transparency, 2023.
[BibTeX] [Link]@inproceedings{258987867, title = {Examining risks of racial biases in NLP tools for child protective services}, author = {{Anjalie Field} and {Amanda Coston} and {Nupoor Gandhi} and {A. Chouldechova} and {Emily Putnam-Hornstein} and {David Steier} and {Yulia Tsvetkov}}, year = 2023, month = {5}, booktitle = {Conference on Fairness, Accountability and Transparency}, url = {https://www.semanticscholar.org/paper/346e4f35a5a81ef893792133ec1fec18f23c1768}, } - Kangfu Mei, Mo Zhou, and Vishal M. Patel, “T1: Scaling Diffusion Probabilistic Fields to High-Resolution on Unified Visual Modalities,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258865247, title = {T1: Scaling Diffusion Probabilistic Fields to High-Resolution on Unified Visual Modalities}, author = {{Kangfu Mei} and {Mo Zhou} and {Vishal M. Patel}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/95dd7685e9b4e6195d80b22167d980be4379da44}, } - A. Favaro, C. Motley, Tianyu Cao, Miguel Iglesias, A. Butala, E. Oh, R. Stevens, J. Villalba, N. Dehak, and L. Moro-Velázquez, “A Multi-Modal Array of Interpretable Features to Evaluate Language and Speech Patterns in Different Neurological Disorders,” in Spoken Language Technology Workshop, 2023.
[BibTeX] [Link]@inproceedings{256353599, title = {A Multi-Modal Array of Interpretable Features to Evaluate Language and Speech Patterns in Different Neurological Disorders}, author = {{A. Favaro} and {C. Motley} and {Tianyu Cao} and {Miguel Iglesias} and {A. Butala} and {E. Oh} and {R. Stevens} and {J. Villalba} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2023, month = {1}, booktitle = {Spoken Language Technology Workshop}, url = {https://www.semanticscholar.org/paper/40eb935374d67b7b9979e0c9333c291d188c472b}, } - Qixing Hu, Yixiong Chen, Junfei Xiao, Shuwen Sun, Jieneng Chen, A. Yuille, and Zongwei Zhou, “Label-Free Liver Tumor Segmentation,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{257766789, title = {Label-Free Liver Tumor Segmentation}, author = {{Qixing Hu} and {Yixiong Chen} and {Junfei Xiao} and {Shuwen Sun} and {Jieneng Chen} and {A. Yuille} and {Zongwei Zhou}}, year = 2023, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/74fc777becc43b9e94c2fb59ed3ee78d212ca01e}, } - Chun-Chieh Chang, Leibny Paola García-Perera, and S. Khudanpur, “Crosslingual Handwritten Text Generation Using GANs,” in ICDAR Workshops, 2023.
[BibTeX] [Link]@inproceedings{261065732, title = {Crosslingual Handwritten Text Generation Using GANs}, author = {{Chun-Chieh Chang} and {Leibny Paola García-Perera} and {S. Khudanpur}}, year = 2023, booktitle = {ICDAR Workshops}, url = {https://www.semanticscholar.org/paper/48abc94b0eaf32dac2573dc5cdbe5dcfa897b7f0}, } - Daniel R. Mendat, Jonah P. Sengupta, Gaspar Tognetti, M. Villemur, P. Pouliquen, Sergio Montano, Kayode A. Sanni, J. Molin, Nishant Zachariah, I. Doxas, and A. Andreou, “A RISC-V Neuromorphic Micro-Controller Unit (vMCU) with Event-Based Physical Interface and Computational Memory for Low-Latency Machine Perception and Intelligence at the Edge,” in International Symposium on Circuits and Systems, 2023.
[BibTeX] [Link]@inproceedings{260003158, title = {A RISC-V Neuromorphic Micro-Controller Unit (vMCU) with Event-Based Physical Interface and Computational Memory for Low-Latency Machine Perception and Intelligence at the Edge}, author = {{Daniel R. Mendat} and {Jonah P. Sengupta} and {Gaspar Tognetti} and {M. Villemur} and {P. Pouliquen} and {Sergio Montano} and {Kayode A. Sanni} and {J. Molin} and {Nishant Zachariah} and {I. Doxas} and {A. Andreou}}, year = 2023, month = {5}, booktitle = {International Symposium on Circuits and Systems}, url = {https://www.semanticscholar.org/paper/0d2f0f6eb40d3be7b97a19315439721cf7ae8469}, } - Neha Verma, Kenton Murray, and Kevin Duh, “Exploring Representational Disparities Between Multilingual and Bilingual Translation Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258841100, title = {Exploring Representational Disparities Between Multilingual and Bilingual Translation Models}, author = {{Neha Verma} and {Kenton Murray} and {Kevin Duh}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/6321d7eec951dd1c6cea44a45f425b774d1b6b26}, } - A. Hussein, Brian Yan, Antonios Anastasopoulos, Shinji Watanabe, and S. Khudanpur, “Enhancing End-to-End Conversational Speech Translation Through Target Language Context Utilization,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263152687, title = {Enhancing End-to-End Conversational Speech Translation Through Target Language Context Utilization}, author = {{A. Hussein} and {Brian Yan} and {Antonios Anastasopoulos} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/c2745e86ecc9bec372690cced53ccfdf44f407f8}, } - Shijie Wu, Ozan Irsoy, Steven Lu, Vadim Dabravolski, Mark Dredze, Sebastian Gehrmann, P. Kambadur, D. Rosenberg, and Gideon Mann, “BloombergGPT: A Large Language Model for Finance,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257833842, title = {BloombergGPT: A Large Language Model for Finance}, author = {{Shijie Wu} and {Ozan Irsoy} and {Steven Lu} and {Vadim Dabravolski} and {Mark Dredze} and {Sebastian Gehrmann} and {P. Kambadur} and {D. Rosenberg} and {Gideon Mann}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/83edcfbb206ddad38a971d605da09390604248ea}, } - Samik Sadhu and H. Hermansky, “Self-supervised Learning with Speech Modulation Dropout,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257687723, title = {Self-supervised Learning with Speech Modulation Dropout}, author = {{Samik Sadhu} and {H. Hermansky}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/eb6358eca5f4ee632f929cb384d07b6a5f04e0ef}, } - Alicia M. Braxton, A. Kiemen, Mia P. Grahn, André Forjaz, Jaanvi Mahesh Babu, Lily Zheng, Li-yu Jiang, H. Cheng, Q. Song, Rebecca Reichel, Sarah Graham, A. Damanakis, Catherine G. Fischer, Stephanie Mou, Cameron Metz, Julie Granger, Xiao-ding Liu, N. Bachmann, Cristina Almagro-Pérez, A. C. Jiang, Jeonghyun Yoo, Bridgette Kim, Scott Du, Eli Foster, Jocelyn Y Hsu, P.A. Rivera, L. Chu, Fengze Liu, N. Niknafs, E. Fishman, A. Yuille, Nicholas J. Roberts, E. Thompson, R. Scharpf, T. Cornish, Y. Jiao, R. Karchin, R. Hruban, Pei-Hsun Wu, D. Wirtz, and L. Wood, “Three-dimensional genomic mapping of human pancreatic tissue reveals striking multifocality and genetic heterogeneity in precancerous lesions,” in bioRxiv, 2023.
[BibTeX] [Link]@inproceedings{256391481, title = {Three-dimensional genomic mapping of human pancreatic tissue reveals striking multifocality and genetic heterogeneity in precancerous lesions}, author = {{Alicia M. Braxton} and {A. Kiemen} and {Mia P. Grahn} and {André Forjaz} and {Jaanvi Mahesh Babu} and {Lily Zheng} and {Li-yu Jiang} and {H. Cheng} and {Q. Song} and {Rebecca Reichel} and {Sarah Graham} and {A. Damanakis} and {Catherine G. Fischer} and {Stephanie Mou} and {Cameron Metz} and {Julie Granger} and {Xiao-ding Liu} and {N. Bachmann} and {Cristina Almagro-Pérez} and {A. C. Jiang} and {Jeonghyun Yoo} and {Bridgette Kim} and {Scott Du} and {Eli Foster} and {Jocelyn Y Hsu} and {P.A. Rivera} and {L. Chu} and {Fengze Liu} and {N. Niknafs} and {E. Fishman} and {A. Yuille} and {Nicholas J. Roberts} and {E. Thompson} and {R. Scharpf} and {T. Cornish} and {Y. Jiao} and {R. Karchin} and {R. Hruban} and {Pei-Hsun Wu} and {D. Wirtz} and {L. Wood}}, year = 2023, month = {1}, booktitle = {bioRxiv}, url = {https://www.semanticscholar.org/paper/8d98352a5fd535de9be9e83fb00ee8ba32fd2761}, } - Wufei Ma, Qihao Liu, Jiahao Wang, Angtian Wang, Yaoyao Liu, Adam Kortylewski, and A. Yuille, “Adding 3D Geometry Control to Diffusion Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259164709, title = {Adding 3D Geometry Control to Diffusion Models}, author = {{Wufei Ma} and {Qihao Liu} and {Jiahao Wang} and {Angtian Wang} and {Yaoyao Liu} and {Adam Kortylewski} and {A. Yuille}}, year = 2023, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/2412b657fbd6f3e366cae47133507a48361d3676}, } - Jeremy Gwinnup and Kevin Duh, “A Survey of Vision-Language Pre-training from the Lens of Multimodal Machine Translation,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259137602, title = {A Survey of Vision-Language Pre-training from the Lens of Multimodal Machine Translation}, author = {{Jeremy Gwinnup} and {Kevin Duh}}, year = 2023, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/c581d2ad3b092a2cc152d0c6f55fd6320f78eb3a}, } - Yixiao Zhang, Xinyi Li, Huimiao Chen, A. Yuille, Yaoyao Liu, and Zongwei Zhou, “Continual Learning for Abdominal Multi-Organ and Tumor Segmentation,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2023.
[BibTeX] [Link]@inproceedings{258999199, title = {Continual Learning for Abdominal Multi-Organ and Tumor Segmentation}, author = {{Yixiao Zhang} and {Xinyi Li} and {Huimiao Chen} and {A. Yuille} and {Yaoyao Liu} and {Zongwei Zhou}}, year = 2023, month = {6}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/7f29e3cb212df207146c567a420999cba6d9fff8}, } - Angtian Wang, Wufei Ma, A. Yuille, and Adam Kortylewski, “Neural Textured Deformable Meshes for Robust Analysis-by-Synthesis,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258999635, title = {Neural Textured Deformable Meshes for Robust Analysis-by-Synthesis}, author = {{Angtian Wang} and {Wufei Ma} and {A. Yuille} and {Adam Kortylewski}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7752104e311751cc786a677f050e2815281039f8}, } - Yuhui Xu, Lingxi Xie, Cihang Xie, Wenrui Dai, Jieru Mei, Siyuan Qiao, Wei Shen, H. Xiong, and A. Yuille, “BNET: Batch Normalization With Enhanced Linear Transformation,” in IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023.
[BibTeX] [Link]@inproceedings{255669634, title = {BNET: Batch Normalization With Enhanced Linear Transformation}, author = {{Yuhui Xu} and {Lingxi Xie} and {Cihang Xie} and {Wenrui Dai} and {Jieru Mei} and {Siyuan Qiao} and {Wei Shen} and {H. Xiong} and {A. Yuille}}, year = 2023, month = {1}, booktitle = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, url = {https://www.semanticscholar.org/paper/edcf374466f791118acf3bbd8430d4fd73e4ea79}, } - Cihan Xiao, Henry Li Xinyuan, Jinyi Yang, Dongji Gao, Matthew Wiesner, Kevin Duh, and S. Khudanpur, “HK-LegiCoST: Leveraging Non-Verbatim Transcripts for Speech Translation,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{259203410, title = {HK-LegiCoST: Leveraging Non-Verbatim Transcripts for Speech Translation}, author = {{Cihan Xiao} and {Henry Li Xinyuan} and {Jinyi Yang} and {Dongji Gao} and {Matthew Wiesner} and {Kevin Duh} and {S. Khudanpur}}, year = 2023, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/74173dec94055d7f4051aa2e80be31ccd2bde596}, } - Kapil D. Katyal, R. Chellappa, Ketul Shah, Arun V. Reddy, Judy Hoffman, William Paul, Rohita Mocharla, D. Handelman, and Celso de Melo, “Leveraging synthetic data for robust gesture recognition,” in Defense + Commercial Sensing, 2023.
[BibTeX] [Link]@inproceedings{258383693, title = {Leveraging synthetic data for robust gesture recognition}, author = {{Kapil D. Katyal} and {R. Chellappa} and {Ketul Shah} and {Arun V. Reddy} and {Judy Hoffman} and {William Paul} and {Rohita Mocharla} and {D. Handelman} and {Celso de Melo}}, year = 2023, month = {6}, booktitle = {Defense + Commercial Sensing}, url = {https://www.semanticscholar.org/paper/922198774621861436721bd923dc0f0028872a84}, } - Jieneng Chen, Yingda Xia, Jiawen Yao, K. Yan, Jianpeng Zhang, Le Lu, Fakai Wang, Bo Zhou, Mingyan Qiu, Qihang Yu, Ming Yuan, Wei Fang, Yuxing Tang, Minfeng Xu, Jian Zhou, Yuqian Zhao, Qifeng Wang, X. Ye, Xiaoli Yin, Yu Shi, Xin Chen, Jingren Zhou, A. Yuille, Zai-De Liu, and Ling Zhang, “CancerUniT: Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans.” 2023.
[BibTeX] [Link]@inproceedings{256389683, title = {CancerUniT: Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans}, author = {{Jieneng Chen} and {Yingda Xia} and {Jiawen Yao} and {K. Yan} and {Jianpeng Zhang} and {Le Lu} and {Fakai Wang} and {Bo Zhou} and {Mingyan Qiu} and {Qihang Yu} and {Ming Yuan} and {Wei Fang} and {Yuxing Tang} and {Minfeng Xu} and {Jian Zhou} and {Yuqian Zhao} and {Qifeng Wang} and {X. Ye} and {Xiaoli Yin} and {Yu Shi} and {Xin Chen} and {Jingren Zhou} and {A. Yuille} and {Zai-De Liu} and {Ling Zhang}}, year = 2023, month = {1}, booktitle = {}, url = {https://www.semanticscholar.org/paper/30be551bfce603f76ba526f047ec4e610b0c2376}, } - Zihao Xiao, A. Yuille, and Yi-Ting Chen, “Learning Road Scene-level Representations via Semantic Region Prediction,” in Conference on Robot Learning, 2023.
[BibTeX] [Link]@inproceedings{255372874, title = {Learning Road Scene-level Representations via Semantic Region Prediction}, author = {{Zihao Xiao} and {A. Yuille} and {Yi-Ting Chen}}, year = 2023, month = {1}, booktitle = {Conference on Robot Learning}, url = {https://www.semanticscholar.org/paper/11b29ca1a235d80a2e55f6eb7711d2aa5785bb8c}, } - A. Kala, E. McCollum, and Mounya Elhilali, “Reference free auscultation quality metric and its trends,” in Biomedical Signal Processing and Control, 2023.
[BibTeX] [Link]@inproceedings{257735707, title = {Reference free auscultation quality metric and its trends}, author = {{A. Kala} and {E. McCollum} and {Mounya Elhilali}}, year = 2023, month = {8}, booktitle = {Biomedical Signal Processing and Control}, url = {https://www.semanticscholar.org/paper/4276e26be8c196ba4b496b4a0acc4102d32c0bd8}, } - K. Shridhar, Harsh Jhamtani, Hao Fang, Benjamin Van Durme, Jason Eisner, and Patrick Xia, “SCREWS: A Modular Framework for Reasoning with Revisions,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{262466051, title = {SCREWS: A Modular Framework for Reasoning with Revisions}, author = {{K. Shridhar} and {Harsh Jhamtani} and {Hao Fang} and {Benjamin Van Durme} and {Jason Eisner} and {Patrick Xia}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/5d5dbba58aaf5b3fa4044cc3ffc71a3fe2b8c654}, } - Miguel Angrick, Shiyu Luo, Qinwan Rabbani, Daniel Candrea, Samyak Shah, Griffin W. Milsap, W. Anderson, C. Gordon, Kathryn R Rosenblatt, L. Clawson, N. Maragakis, F. Tenore, M. Fifer, H. Hermansky, N. Ramsey, and N. Crone, “Online speech synthesis using a chronically implanted brain-computer interface in an individual with ALS,” in medRxiv, 2023.
[BibTeX] [Link]@inproceedings{259302532, title = {Online speech synthesis using a chronically implanted brain-computer interface in an individual with ALS}, author = {{Miguel Angrick} and {Shiyu Luo} and {Qinwan Rabbani} and {Daniel Candrea} and {Samyak Shah} and {Griffin W. Milsap} and {W. Anderson} and {C. Gordon} and {Kathryn R Rosenblatt} and {L. Clawson} and {N. Maragakis} and {F. Tenore} and {M. Fifer} and {H. Hermansky} and {N. Ramsey} and {N. Crone}}, year = 2023, month = {7}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/3991788eee23646956065f50303a62379c1150dd}, } - Z. Smith, N. Hoekstra, T. Mvalo, I. McLane, A. Kala, M. Chiume, C. Verwey, D. Olson, C. Buck, J. Mulindwa, E. Fitzgerald, M. Chagomerana, Mounya Elhilali, M. Hosseinipour, and E. McCollum, “Evaluation of a Novel Digital Stethoscope Prototype in a Low-resource Setting: Expert Listening Panel Agreement With Conventional Auscultation in Hospitalized Malawian Children With Severe Pneumonia,” in C25. OPPORTUNITIES AND ADVANCES IN PEDIATRIC GLOBAL HEALTH, 2023.
[BibTeX] [Link]@inproceedings{258444556, title = {Evaluation of a Novel Digital Stethoscope Prototype in a Low-resource Setting: Expert Listening Panel Agreement With Conventional Auscultation in Hospitalized Malawian Children With Severe Pneumonia}, author = {{Z. Smith} and {N. Hoekstra} and {T. Mvalo} and {I. McLane} and {A. Kala} and {M. Chiume} and {C. Verwey} and {D. Olson} and {C. Buck} and {J. Mulindwa} and {E. Fitzgerald} and {M. Chagomerana} and {Mounya Elhilali} and {M. Hosseinipour} and {E. McCollum}}, year = 2023, month = {5}, booktitle = {C25. OPPORTUNITIES AND ADVANCES IN PEDIATRIC GLOBAL HEALTH}, url = {https://www.semanticscholar.org/paper/00ff74d263d80498ea78cca8850c565b66057476}, } - Paul McNamee and Kevin Duh, “An Extensive Exploration of Back-Translation in 60 Languages,” in Annual Meeting of the Association for Computational Linguistics, 2023.
[BibTeX] [Link]@inproceedings{259859140, title = {An Extensive Exploration of Back-Translation in 60 Languages}, author = {{Paul McNamee} and {Kevin Duh}}, year = 2023, booktitle = {Annual Meeting of the Association for Computational Linguistics}, url = {https://www.semanticscholar.org/paper/3b1cea929fb0a44886ed654c9ca88a9df959f371}, } - Jiacong Xu, Yi Zhang, Jia-Xiong Peng, Wufei Ma, Artur Jesslen, Pengliang Ji, Qixing Hu, Jiehua Zhang, Qihao Liu, Jiahao Wang, Wei Ji, Chen Wang, Xiaoding Yuan, Prakhar Kaushik, Guofeng Zhang, Jie Liu, Yushan Xie, Yawen Cui, A. Yuille, and Adam Kortylewski, “Animal3D: A Comprehensive Dataset of 3D Animal Pose and Shape,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{261076475, title = {Animal3D: A Comprehensive Dataset of 3D Animal Pose and Shape}, author = {{Jiacong Xu} and {Yi Zhang} and {Jia-Xiong Peng} and {Wufei Ma} and {Artur Jesslen} and {Pengliang Ji} and {Qixing Hu} and {Jiehua Zhang} and {Qihao Liu} and {Jiahao Wang} and {Wei Ji} and {Chen Wang} and {Xiaoding Yuan} and {Prakhar Kaushik} and {Guofeng Zhang} and {Jie Liu} and {Yushan Xie} and {Yawen Cui} and {A. Yuille} and {Adam Kortylewski}}, year = 2023, month = {8}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/cf6f0b77e006083e74d5f08bae59bd207d0e4ac6}, } - Bingchen Zhao, Jiahao Wang, Wufei Ma, Artur Jesslen, Si-Jia Yang, Shaozuo Yu, O. Zendel, C. Theobalt, A. Yuille, and Adam Kortylewski, “OOD-CV-v2: An extended Benchmark for Robustness to Out-of-Distribution Shifts of Individual Nuisances in Natural Images,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258236142, title = {OOD-CV-v2: An extended Benchmark for Robustness to Out-of-Distribution Shifts of Individual Nuisances in Natural Images}, author = {{Bingchen Zhao} and {Jiahao Wang} and {Wufei Ma} and {Artur Jesslen} and {Si-Jia Yang} and {Shaozuo Yu} and {O. Zendel} and {C. Theobalt} and {A. Yuille} and {Adam Kortylewski}}, year = 2023, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/f5cd9b3f48e81e1a91923ef423765edeb9bdd50e}, } - Jeya Maria Jose Valanarasu, Rahul Garg, Andeep S. Toor, Xin Tong, Weijuan Xi, Andreas Lugmayr, Vishal M. Patel, and A. Menini, “ReBotNet: Fast Real-time Video Enhancement,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257687311, title = {ReBotNet: Fast Real-time Video Enhancement}, author = {{Jeya Maria Jose Valanarasu} and {Rahul Garg} and {Andeep S. Toor} and {Xin Tong} and {Weijuan Xi} and {Andreas Lugmayr} and {Vishal M. Patel} and {A. Menini}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/15c2b3ecdf1b9af2f94a2b106fddcfc89cb336cb}, } - Shao-Yuan Lo, Poojan Oza, Sumanth Chennupati, Alejandro Galindo, and Vishal M. Patel, “Spatio-Temporal Pixel-Level Contrastive Learning-based Source-Free Domain Adaptation for Video Semantic Segmentation,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{257766699, title = {Spatio-Temporal Pixel-Level Contrastive Learning-based Source-Free Domain Adaptation for Video Semantic Segmentation}, author = {{Shao-Yuan Lo} and {Poojan Oza} and {Sumanth Chennupati} and {Alejandro Galindo} and {Vishal M. Patel}}, year = 2023, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/47cd9158e970329355a575ed992d4452ac498784}, } - Lauren M. Sanders, Ryan T. Scott, Jason H. Yang, Amina Ann Qutub, Héctor García Martín, D. Berrios, Jaden J. A. Hastings, J. Rask, Graham Mackintosh, A. Hoarfrost, Stuart Chalk, John Kalantari, K. Khezeli, E. Antonsen, Joel Babdor, Richard Barker, S. Baranzini, Afshin Beheshti, Guillermo M. Delgado-Aparicio, B. Glicksberg, Casey S. Greene, Melissa Haendel, Arif A. Hamid, P. Heller, Daniel Jamieson, K. Jarvis, Svetlana V. Komarova, M. Komorowski, Prachi Kothiyal, A. Mahabal, U. Manor, Christopher E. Mason, Mona Matar, G. Mias, Jack M. Miller, J. Myers, Charlotte A. Nelson, Jonathan Oribello, Seung-min Park, P. Parsons-Wingerter, R. K. Prabhu, R. Reynolds, Amanda M. Saravia-Butler, S. Saria, A. Sawyer, N. Singh, M. Snyder, Frank Soboczenski, Karthik Soman, C. Theriot, David Van Valen, K. Venkateswaran, L. Warren, Liz Worthey, M. Zitnik, and S. Costes, “Biological research and self-driving labs in deep space supported by artificial intelligence,” in Nature Machine Intelligence, 2023.
[BibTeX] [Link]@inproceedings{257697867, title = {Biological research and self-driving labs in deep space supported by artificial intelligence}, author = {{Lauren M. Sanders} and {Ryan T. Scott} and {Jason H. Yang} and {Amina Ann Qutub} and {Héctor García Martín} and {D. Berrios} and {Jaden J. A. Hastings} and {J. Rask} and {Graham Mackintosh} and {A. Hoarfrost} and {Stuart Chalk} and {John Kalantari} and {K. Khezeli} and {E. Antonsen} and {Joel Babdor} and {Richard Barker} and {S. Baranzini} and {Afshin Beheshti} and {Guillermo M. Delgado-Aparicio} and {B. Glicksberg} and {Casey S. Greene} and {Melissa Haendel} and {Arif A. Hamid} and {P. Heller} and {Daniel Jamieson} and {K. Jarvis} and {Svetlana V. Komarova} and {M. Komorowski} and {Prachi Kothiyal} and {A. Mahabal} and {U. Manor} and {Christopher E. Mason} and {Mona Matar} and {G. Mias} and {Jack M. Miller} and {J. Myers} and {Charlotte A. Nelson} and {Jonathan Oribello} and {Seung-min Park} and {P. Parsons-Wingerter} and {R. K. Prabhu} and {R. Reynolds} and {Amanda M. Saravia-Butler} and {S. Saria} and {A. Sawyer} and {N. Singh} and {M. Snyder} and {Frank Soboczenski} and {Karthik Soman} and {C. Theriot} and {David Van Valen} and {K. Venkateswaran} and {L. Warren} and {Liz Worthey} and {M. Zitnik} and {S. Costes}}, year = 2023, month = {3}, booktitle = {Nature Machine Intelligence}, url = {https://www.semanticscholar.org/paper/880e7f45c1952189e350545dd98a73ef47465cba}, } - Marc Marone and Benjamin Van Durme, “Data Portraits: Recording Foundation Model Training Data,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257378087, title = {Data Portraits: Recording Foundation Model Training Data}, author = {{Marc Marone} and {Benjamin Van Durme}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/572b92972eff7501ca2b109b8998cdcb69aa1958}, } - Huali Xu, S. Zhi, Shuzhou Sun, Vishal M. Patel, and Li Liu, “Deep Learning for Cross-Domain Few-Shot Visual Recognition: A Survey,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257532548, title = {Deep Learning for Cross-Domain Few-Shot Visual Recognition: A Survey}, author = {{Huali Xu} and {S. Zhi} and {Shuzhou Sun} and {Vishal M. Patel} and {Li Liu}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/095138d9207da38bce4914c569e2f312927213b5}, } - Haoran Xu, Weiting Tan, Shuyue Stella Li, Yunmo Chen, Benjamin Van Durme, Philipp Koehn, and Kenton Murray, “Condensing Multilingual Knowledge with Lightweight Language-Specific Modules,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258841086, title = {Condensing Multilingual Knowledge with Lightweight Language-Specific Modules}, author = {{Haoran Xu} and {Weiting Tan} and {Shuyue Stella Li} and {Yunmo Chen} and {Benjamin Van Durme} and {Philipp Koehn} and {Kenton Murray}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/cf99558d35891d35ba5429cfce5ca2e2ae383b71}, } - J. Brugts, S. Radhoe, P. R. Clephas, Dilan Aydin, M. Gent, M. Szymanski, M. Rienstra, Mieke H van den Heuvel, Carlos A da Fonseca, G. Linssen, C. Borleffs, E. Boersma, F. Asselbergs, A. Mosterd, H. Rocca, R. A. Boer, M. Emans, S. Beeres, L. Heerebeek, C. Kirchhof, J. Ramshorst, R. Spee, T. Smilde, M. V. Eck, E. Kaplan, R. Hazeleger, R. Tukkie, M. Feenema, W. Kok, V. V. Halm, M. L. Handoko, R. Kimmenade, Matt Post, N. V. Mieghem, and O. Manintveld, “Remote haemodynamic monitoring of pulmonary artery pressures in patients with chronic heart failure (MONITOR-HF): a randomised clinical trial,” in The Lancet, 2023.
[BibTeX] [Link]@inproceedings{258808787, title = {Remote haemodynamic monitoring of pulmonary artery pressures in patients with chronic heart failure (MONITOR-HF): a randomised clinical trial}, author = {{J. Brugts} and {S. Radhoe} and {P. R. Clephas} and {Dilan Aydin} and {M. Gent} and {M. Szymanski} and {M. Rienstra} and {Mieke H van den Heuvel} and {Carlos A da Fonseca} and {G. Linssen} and {C. Borleffs} and {E. Boersma} and {F. Asselbergs} and {A. Mosterd} and {H. Rocca} and {R. A. Boer} and {M. Emans} and {S. Beeres} and {L. Heerebeek} and {C. Kirchhof} and {J. Ramshorst} and {R. Spee} and {T. Smilde} and {M. V. Eck} and {E. Kaplan} and {R. Hazeleger} and {R. Tukkie} and {M. Feenema} and {W. Kok} and {V. V. Halm} and {M. L. Handoko} and {R. Kimmenade} and {Matt Post} and {N. V. Mieghem} and {O. Manintveld}}, year = 2023, month = {5}, booktitle = {The Lancet}, url = {https://www.semanticscholar.org/paper/80da514a8c411b22bd786a69f3ca62ae1d323ff1}, } - Vikas Raunak, Tom Kocmi, and Matt Post, “SLIDE: Reference-free Evaluation for Machine Translation using a Sliding Document Window,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{262043615, title = {SLIDE: Reference-free Evaluation for Machine Translation using a Sliding Document Window}, author = {{Vikas Raunak} and {Tom Kocmi} and {Matt Post}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/eea9f4ad5af30667535df7aa954c63f7cbc9f9da}, } - Lingfeng Shen, Sihao Chen, Linfeng Song, Lifeng Jin, Baolin Peng, Haitao Mi, Daniel Khashabi, and Dong Yu, “The Trickle-down Impact of Reward (In-)consistency on RLHF,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{263134276, title = {The Trickle-down Impact of Reward (In-)consistency on RLHF}, author = {{Lingfeng Shen} and {Sihao Chen} and {Linfeng Song} and {Lifeng Jin} and {Baolin Peng} and {Haitao Mi} and {Daniel Khashabi} and {Dong Yu}}, year = 2023, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/541b66bad4a9bf9b7fd97f13f94ab9061c7c0c47}, } - Jiahao Yang, Wufei Ma, Angtian Wang, Xiaoding Yuan, A. Yuille, and Adam Kortylewski, “Robust Category-Level 3D Pose Estimation from Synthetic Data,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258887517, title = {Robust Category-Level 3D Pose Estimation from Synthetic Data}, author = {{Jiahao Yang} and {Wufei Ma} and {Angtian Wang} and {Xiaoding Yuan} and {A. Yuille} and {Adam Kortylewski}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/55aa226650e6eeed51e181195b7b7a9b87102bc5}, } - Gerardo Flores, George H. Chen, T. Pollard, Ayah Zirikly, Michael C. Hughes, Tasmie Sarker, Joyce Ho, and Tristan Naumann, “Conference on Health, Inference, and Learning, CHIL 2023, Broad Institute of MIT and Harvard (Merkin Building), 415 Main Street, Cambridge, MA, USA,” in ACM Conference on Health, Inference, and Learning, 2023.
[BibTeX] [Link]@inproceedings{248397783, title = {Conference on Health, Inference, and Learning, CHIL 2023, Broad Institute of MIT and Harvard (Merkin Building), 415 Main Street, Cambridge, MA, USA}, author = {{Gerardo Flores} and {George H. Chen} and {T. Pollard} and {Ayah Zirikly} and {Michael C. Hughes} and {Tasmie Sarker} and {Joyce Ho} and {Tristan Naumann}}, year = 2023, booktitle = {ACM Conference on Health, Inference, and Learning}, url = {https://www.semanticscholar.org/paper/1cfe0feed33d2452f951afa7304d017131dc4520}, } - Samuel Barham, Orion Weller, Michelle Yuan, Kenton Murray, M. Yarmohammadi, Zhengping Jiang, Siddharth Vashishtha, Alexander Martin, Anqi Liu, Aaron Steven White, Jordan L. Boyd-Graber, and Benjamin Van Durme, “MegaWika: Millions of reports and their sources across 50 diverse languages,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{259924959, title = {MegaWika: Millions of reports and their sources across 50 diverse languages}, author = {{Samuel Barham} and {Orion Weller} and {Michelle Yuan} and {Kenton Murray} and {M. Yarmohammadi} and {Zhengping Jiang} and {Siddharth Vashishtha} and {Alexander Martin} and {Anqi Liu} and {Aaron Steven White} and {Jordan L. Boyd-Graber} and {Benjamin Van Durme}}, year = 2023, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/c8e2b9f68bd585759d31741193516f58b5619584}, } - Wei Ji, Jingjing Li, Cheng Bian, Zongwei Zhou, Jiaying Zhao, A. Yuille, and Li Cheng, “Multispectral Video Semantic Segmentation: A Benchmark Dataset and Baseline,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{259840157, title = {Multispectral Video Semantic Segmentation: A Benchmark Dataset and Baseline}, author = {{Wei Ji} and {Jingjing Li} and {Cheng Bian} and {Zongwei Zhou} and {Jiaying Zhao} and {A. Yuille} and {Li Cheng}}, year = 2023, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/fa0215c456d1957e131e43702e50e4c8f3e9477d}, } - Deepti Hegde, Jeya Maria Jose Valanarasu, and Vishal M. Patel, “CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{257632366, title = {CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition}, author = {{Deepti Hegde} and {Jeya Maria Jose Valanarasu} and {Vishal M. Patel}}, year = 2023, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b460a263abec8b1aaa039963be9b371a581e7b21}, } - Manoj Jain, Salil Bhargava, R. Arora, Rajendra Joshi, Ravinder Kumar, Deepak Saxena, Kiran Rade, and Rebecca Martin, “Using a Quality Improvement Tool, Plan-Do-Study-Act Cycle, to Boost TB Notification in India post-Covid-19 Pandemic,” in Indian Journal of Tuberculosis, 2023.
[BibTeX] [Link]@inproceedings{262185166, title = {Using a Quality Improvement Tool, Plan-Do-Study-Act Cycle, to Boost TB Notification in India post-Covid-19 Pandemic}, author = {{Manoj Jain} and {Salil Bhargava} and {R. Arora} and {Rajendra Joshi} and {Ravinder Kumar} and {Deepak Saxena} and {Kiran Rade} and {Rebecca Martin}}, year = 2023, month = {9}, booktitle = {Indian Journal of Tuberculosis}, url = {https://www.semanticscholar.org/paper/acbaffb72d4c3bd7c9a12d6c756a4a207dea3703}, } - Drew Prinster, S. Saria, and Anqi Liu, “Efficient Approximate Predictive Inference Under Feedback Covariate Shift with Influence Functions,” in International Symposium on Conformal and Probabilistic Prediction with Applications, 2023.
[BibTeX] [Link]@inproceedings{262070652, title = {Efficient Approximate Predictive Inference Under Feedback Covariate Shift with Influence Functions}, author = {{Drew Prinster} and {S. Saria} and {Anqi Liu}}, year = 2023, booktitle = {International Symposium on Conformal and Probabilistic Prediction with Applications}, url = {https://www.semanticscholar.org/paper/ea62b78ec46e7ca0ad4dd5337cd87bb27ab0ec06}, } - Tianyu Cao, L. Moro-Velázquez, Piotr Żelasko, J. Villalba, and N. Dehak, “Vsameter: Evaluation of a New Open-Source Tool to Measure Vowel Space Area and Related Metrics,” in Spoken Language Technology Workshop, 2023.
[BibTeX] [Link]@inproceedings{256356339, title = {Vsameter: Evaluation of a New Open-Source Tool to Measure Vowel Space Area and Related Metrics}, author = {{Tianyu Cao} and {L. Moro-Velázquez} and {Piotr Żelasko} and {J. Villalba} and {N. Dehak}}, year = 2023, month = {1}, booktitle = {Spoken Language Technology Workshop}, url = {https://www.semanticscholar.org/paper/dd3d00bf410d95d15569443387082da13a2462c4}, } - Qihao Liu, Adam Kortylewski, Yutong Bai, Song Bai, and A. Yuille, “Intriguing Properties of Text-guided Diffusion Models,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258999203, title = {Intriguing Properties of Text-guided Diffusion Models}, author = {{Qihao Liu} and {Adam Kortylewski} and {Yutong Bai} and {Song Bai} and {A. Yuille}}, year = 2023, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/5a7a7655a7cc25e78a20892315902dceaaba3bf4}, } - Denis Newman-Griffis, Bart Desmet, Ayah Zirikly, Suzanne Tamang, and Chih-Hung Chang, “Editorial: Artificial intelligence for human function and disability,” in Frontiers in Digital Health, 2023.
[BibTeX] [Link]@inproceedings{261637808, title = {Editorial: Artificial intelligence for human function and disability}, author = {{Denis Newman-Griffis} and {Bart Desmet} and {Ayah Zirikly} and {Suzanne Tamang} and {Chih-Hung Chang}}, year = 2023, month = {9}, booktitle = {Frontiers in Digital Health}, url = {https://www.semanticscholar.org/paper/251b6ab5f8aa64447bcab84b2078a7198afc4ac3}, } - Saurabh Kataria, J. Villalba, Laureano Moro-Vel’azquez, Thomas Thebaud, and N. Dehak, “Self-FiLM: Conditioning GANs with self-supervised representations for bandwidth extension based speaker recognition,” in Interspeech, 2023.
[BibTeX] [Link]@inproceedings{257378503, title = {Self-FiLM: Conditioning GANs with self-supervised representations for bandwidth extension based speaker recognition}, author = {{Saurabh Kataria} and {J. Villalba} and {Laureano Moro-Vel'azquez} and {Thomas Thebaud} and {N. Dehak}}, year = 2023, month = {3}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/03e266795339008e9366daabfd2a2db2fbd51151}, } - Ryan T. Scott, Lauren M. Sanders, E. Antonsen, Jaden J. A. Hastings, Seung-min Park, Graham Mackintosh, R. Reynolds, A. Hoarfrost, A. Sawyer, Casey S. Greene, Benjamin S. Glicksberg, C. Theriot, D. Berrios, Jack Miller, Joel Babdor, Richard Barker, S. Baranzini, Afshin Beheshti, Stuart Chalk, Guillermo M. Delgado-Aparicio, Melissa Haendel, Arif A. Hamid, P. Heller, Daniel Jamieson, K. Jarvis, John Kalantari, K. Khezeli, Svetlana V. Komarova, M. Komorowski, Prachi Kothiyal, A. Mahabal, U. Manor, Héctor García Martín, Christopher E. Mason, Mona Matar, G. Mias, J. Myers, Charlotte A. Nelson, Jonathan Oribello, P. Parsons-Wingerter, R. K. Prabhu, A. Qutub, J. Rask, Amanda M. Saravia-Butler, S. Saria, N. Singh, M. Snyder, Frank Soboczenski, Karthik Soman, David Van Valen, K. Venkateswaran, L. Warren, Liz Worthey, Jason H. Yang, M. Zitnik, and S. Costes, “Biomonitoring and precision health in deep space supported by artificial intelligence,” in Nature Machine Intelligence, 2023.
[BibTeX] [Link]@inproceedings{257698009, title = {Biomonitoring and precision health in deep space supported by artificial intelligence}, author = {{Ryan T. Scott} and {Lauren M. Sanders} and {E. Antonsen} and {Jaden J. A. Hastings} and {Seung-min Park} and {Graham Mackintosh} and {R. Reynolds} and {A. Hoarfrost} and {A. Sawyer} and {Casey S. Greene} and {Benjamin S. Glicksberg} and {C. Theriot} and {D. Berrios} and {Jack Miller} and {Joel Babdor} and {Richard Barker} and {S. Baranzini} and {Afshin Beheshti} and {Stuart Chalk} and {Guillermo M. Delgado-Aparicio} and {Melissa Haendel} and {Arif A. Hamid} and {P. Heller} and {Daniel Jamieson} and {K. Jarvis} and {John Kalantari} and {K. Khezeli} and {Svetlana V. Komarova} and {M. Komorowski} and {Prachi Kothiyal} and {A. Mahabal} and {U. Manor} and {Héctor García Martín} and {Christopher E. Mason} and {Mona Matar} and {G. Mias} and {J. Myers} and {Charlotte A. Nelson} and {Jonathan Oribello} and {P. Parsons-Wingerter} and {R. K. Prabhu} and {A. Qutub} and {J. Rask} and {Amanda M. Saravia-Butler} and {S. Saria} and {N. Singh} and {M. Snyder} and {Frank Soboczenski} and {Karthik Soman} and {David Van Valen} and {K. Venkateswaran} and {L. Warren} and {Liz Worthey} and {Jason H. Yang} and {M. Zitnik} and {S. Costes}}, year = 2023, month = {3}, booktitle = {Nature Machine Intelligence}, url = {https://www.semanticscholar.org/paper/275a42c374d6381406a5da16dfa52fa939817a15}, } - Qihao Liu, Junfeng Wu, Yi Jiang, Xiang Bai, A. Yuille, and S. Bai, “InstMove: Instance Motion for Object-centric Video Segmentation,” in Computer Vision and Pattern Recognition, 2023.
[BibTeX] [Link]@inproceedings{257505032, title = {InstMove: Instance Motion for Object-centric Video Segmentation}, author = {{Qihao Liu} and {Junfeng Wu} and {Yi Jiang} and {Xiang Bai} and {A. Yuille} and {S. Bai}}, year = 2023, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/0e60c1229d7963b605b83cb10a90ed6a8cf79149}, } - Elizabeth Salesky, Neha Verma, Philipp Koehn, and Matt Post, “Multilingual Pixel Representations for Translation and Effective Cross-lingual Transfer.” 2023.
[BibTeX] [Link]@inproceedings{258841479, title = {Multilingual Pixel Representations for Translation and Effective Cross-lingual Transfer}, author = {{Elizabeth Salesky} and {Neha Verma} and {Philipp Koehn} and {Matt Post}}, year = 2023, month = {5}, booktitle = {}, url = {https://www.semanticscholar.org/paper/c60736e61f8961ec535ecfdc6f0398925d34d0b8}, } - Harsh Jhamtani, Hao Fang, Patrick Xia, Eran Levy, Jacob Andreas, and Benjamin Van Durme, “Natural Language Decomposition and Interpretation of Complex Utterances,” in arXiv.org, 2023.
[BibTeX] [Link]@inproceedings{258686491, title = {Natural Language Decomposition and Interpretation of Complex Utterances}, author = {{Harsh Jhamtani} and {Hao Fang} and {Patrick Xia} and {Eran Levy} and {Jacob Andreas} and {Benjamin Van Durme}}, year = 2023, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/68040213e9a83408cdc491ed3e235b52b537eed1}, } - Enayat Ullah and R. Arora, “From Adaptive Query Release to Machine Unlearning,” in International Conference on Machine Learning, 2023.
[BibTeX] [Link]@inproceedings{260091661, title = {From Adaptive Query Release to Machine Unlearning}, author = {{Enayat Ullah} and {R. Arora}}, year = 2023, month = {7}, booktitle = {International Conference on Machine Learning}, url = {https://www.semanticscholar.org/paper/ea3eff68041f3a22b984578e8da8533aa3f766de}, } - D. Li, A. Butala, T. Meyer, E. Oh, C. Motley, L. Moro-Velázquez, and N. Dehak, “Automating analysis of eye movement and feature extraction for different neurodegenerative disorders,” in medRxiv, 2023.
[BibTeX] [Link]@inproceedings{258997975, title = {Automating analysis of eye movement and feature extraction for different neurodegenerative disorders}, author = {{D. Li} and {A. Butala} and {T. Meyer} and {E. Oh} and {C. Motley} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2023, month = {6}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/2b5f1cfc2b507561bd463b0a5ac14fd92d75dc50}, } - A. Favaro, Seneca Motley, Quincy M. Samus, A. Butala, N. Dehak, Esther S. Oh, and L. Moro-Velázquez, “Artificial Intelligence Tools to Evaluate Language and Speech Patterns in Alzheimer’s Disease,” in Alzheimer’s & Dementia, 2022.
[BibTeX] [Link]@inproceedings{254880773, title = {Artificial Intelligence Tools to Evaluate Language and Speech Patterns in Alzheimer's Disease}, author = {{A. Favaro} and {Seneca Motley} and {Quincy M. Samus} and {A. Butala} and {N. Dehak} and {Esther S. Oh} and {L. Moro-Velázquez}}, year = 2022, month = {12}, booktitle = {Alzheimer's & Dementia}, url = {https://www.semanticscholar.org/paper/e8f74514d4b195230ddd7dd6b60cabbc7ed240b1}, } - Y. Feng, P. Xia, B. Van Durme, and J. Sedoc, “Automatic Document Selection for Efficient Encoder Pretraining,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 9522–9530. doi:10.18653/v1/2022.emnlp-main.647
[BibTeX] [Abstract] [Link]
Building pretrained language models is considered expensive and data-intensive, but must we increase dataset size to achieve better performance? We propose an alternative to larger training sets by automatically identifying smaller yet domain-representative subsets. We extend Cynical Data Selection, a statistical sentence scoring method that conditions on a representative target domain corpus. As an example, we treat the OntoNotes corpus as a target domain and pretrain a RoBERTa-like encoder from a cynically selected subset of the Pile. On both perplexity and across several downstream tasks in the target domain, it consistently outperforms random selection with 20x less data, 3x fewer training iterations, and 2x less estimated cloud compute cost, validating the recipe of automatic document selection for LM pretraining.
@inproceedings{feng-etal-2022-automatic, title = "Automatic Document Selection for Efficient Encoder Pretraining", author = "Feng, Yukun and Xia, Patrick and Van Durme, Benjamin and Sedoc, Jo{\~a}o", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.647", doi = "10.18653/v1/2022.emnlp-main.647", pages = "9522--9530", abstract = "Building pretrained language models is considered expensive and data-intensive, but must we increase dataset size to achieve better performance? We propose an alternative to larger training sets by automatically identifying smaller yet domain-representative subsets. We extend Cynical Data Selection, a statistical sentence scoring method that conditions on a representative target domain corpus. As an example, we treat the OntoNotes corpus as a target domain and pretrain a RoBERTa-like encoder from a cynically selected subset of the Pile. On both perplexity and across several downstream tasks in the target domain, it consistently outperforms random selection with 20x less data, 3x fewer training iterations, and 2x less estimated cloud compute cost, validating the recipe of automatic document selection for LM pretraining.", } - Kangfu Mei and Vishal M. Patel, “VIDM: Video Implicit Diffusion Models,” in AAAI Conference on Artificial Intelligence, 2022.
[BibTeX] [Link]@inproceedings{254125713, title = {VIDM: Video Implicit Diffusion Models}, author = {{Kangfu Mei} and {Vishal M. Patel}}, year = 2022, month = {12}, booktitle = {AAAI Conference on Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/13c7b29a100f67d285eb3625c160d06882d4c092}, } - Hongru Zhu, Yijun Ge, Alexander Bratch, A. Yuille, Kendrick Norris Kay, and D. Kersten, “Distributed representations of natural body pose in visual cortex,” in Journal of Vision, 2022.
[BibTeX] [Link]@inproceedings{254767215, title = {Distributed representations of natural body pose in visual cortex}, author = {{Hongru Zhu} and {Yijun Ge} and {Alexander Bratch} and {A. Yuille} and {Kendrick Norris Kay} and {D. Kersten}}, year = 2022, month = {12}, booktitle = {Journal of Vision}, url = {https://www.semanticscholar.org/paper/0f737f04ade2ef8f4a360dc42296476a55fa49d3}, } - Kangfu Mei, Nithin Gopalakrishnan Nair, and Vishal M. Patel, “Bi-Noising Diffusion: Towards Conditional Diffusion Models with Generative Restoration Priors,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254636282, title = {Bi-Noising Diffusion: Towards Conditional Diffusion Models with Generative Restoration Priors}, author = {{Kangfu Mei} and {Nithin Gopalakrishnan Nair} and {Vishal M. Patel}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/e5fec7e9103c9edbc4c6b4bb1a47e53593c667bb}, } - D. D. Kairamkonda, P. S. Mandaleeka, A. Favaro, C. Motley, A. Butala, E. Oh, R. Stevens, N. Dehak, and L. Moro-Velázquez, “Analysis of Interpretable Handwriting Features to Evaluate Motoric Patterns in Different Neurodegenerative Diseases,” in IEEE Signal Processing in Medicine and Biology Symposium, 2022.
[BibTeX] [Link]@inproceedings{256034700, title = {Analysis of Interpretable Handwriting Features to Evaluate Motoric Patterns in Different Neurodegenerative Diseases}, author = {{D. D. Kairamkonda} and {P. S. Mandaleeka} and {A. Favaro} and {C. Motley} and {A. Butala} and {E. Oh} and {R. Stevens} and {N. Dehak} and {L. Moro-Velázquez}}, year = 2022, month = {12}, booktitle = {IEEE Signal Processing in Medicine and Biology Symposium}, url = {https://www.semanticscholar.org/paper/d10f7b6ab049a92c19e1d9c7792063e85ce60d22}, } - F. Casacuberta, G. Foster, G. Huang, P. Koehn, G. Kovacs, L. Liu, S. Shi, T. Watanabe, and C. Zong, “Findings of the Word-Level AutoCompletion Shared Task in WMT 2022,” in Proceedings of the Seventh Conference on Machine Translation (WMT), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 812–820.
[BibTeX] [Abstract] [Link]
Recent years have witnessed rapid advancements in machine translation, but the state-of-the-art machine translation system still can not satisfy the high requirements in some rigorous translation scenarios. Computer-aided translation (CAT) provides a promising solution to yield a high-quality translation with a guarantee. Unfortunately, due to the lack of popular benchmarks, the research on CAT is not well developed compared with machine translation. In this year, we hold a new shared task called Word-level AutoCompletion (WLAC) for CAT in WMT. Specifically, we introduce some resources to train a WLAC model, and particularly we collect data from CAT systems as a part of test data for this shared task. In addition, we employ both automatic and human evaluations to measure the performance of the submitted systems, and our final evaluation results reveal some findings for the WLAC task.
@inproceedings{casacuberta-etal-2022-findings, title = "Findings of the Word-Level {A}uto{C}ompletion Shared Task in {WMT} 2022", author = "Casacuberta, Francisco and Foster, George and Huang, Guoping and Koehn, Philipp and Kovacs, Geza and Liu, Lemao and Shi, Shuming and Watanabe, Taro and Zong, Chengqing", editor = {Koehn, Philipp and Barrault, Lo{\"\i}c and Bojar, Ond{\v{r}}ej and Bougares, Fethi and Chatterjee, Rajen and Costa-juss{\`a}, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Jimeno Yepes, Antonio and Kocmi, Tom and Martins, Andr{\'e} and Morishita, Makoto and Monz, Christof and Nagata, Masaaki and Nakazawa, Toshiaki and Negri, Matteo and N{\'e}v{\'e}ol, Aur{\'e}lie and Neves, Mariana and Popel, Martin and Turchi, Marco and Zampieri, Marcos}, booktitle = "Proceedings of the Seventh Conference on Machine Translation (WMT)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wmt-1.75", pages = "812--820", abstract = "Recent years have witnessed rapid advancements in machine translation, but the state-of-the-art machine translation system still can not satisfy the high requirements in some rigorous translation scenarios. Computer-aided translation (CAT) provides a promising solution to yield a high-quality translation with a guarantee. Unfortunately, due to the lack of popular benchmarks, the research on CAT is not well developed compared with machine translation. In this year, we hold a new shared task called Word-level AutoCompletion (WLAC) for CAT in WMT. Specifically, we introduce some resources to train a WLAC model, and particularly we collect data from CAT systems as a part of test data for this shared task. In addition, we employ both automatic and human evaluations to measure the performance of the submitted systems, and our final evaluation results reveal some findings for the WLAC task.", } - Trevor Meyer, L. Moro-Velázquez, Seneca Motley, A. Butala, Ashley M Paul, Quincy M. Samus, Pedro P. Irazoqui, N. Dehak, and Esther S. Oh, “Automatic Extraction of Oculographic Signals as Digital Biomarkers for Alzheimer’s Disease,” in Alzheimer’s & Dementia, 2022.
[BibTeX] [Link]@inproceedings{254879636, title = {Automatic Extraction of Oculographic Signals as Digital Biomarkers for Alzheimer's Disease}, author = {{Trevor Meyer} and {L. Moro-Velázquez} and {Seneca Motley} and {A. Butala} and {Ashley M Paul} and {Quincy M. Samus} and {Pedro P. Irazoqui} and {N. Dehak} and {Esther S. Oh}}, year = 2022, month = {12}, booktitle = {Alzheimer's & Dementia}, url = {https://www.semanticscholar.org/paper/e5a0988cdd73b981611be9fe06e0b7328ff1c0d0}, } - Si-Jia Yang, Longlong Jing, Junfei Xiao, Hang Zhao, A. Yuille, and Yingwei Li, “AsyInst: Asymmetric Affinity with DepthGrad and Color for Box-Supervised Instance Segmentation,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254366617, title = {AsyInst: Asymmetric Affinity with DepthGrad and Color for Box-Supervised Instance Segmentation}, author = {{Si-Jia Yang} and {Longlong Jing} and {Junfei Xiao} and {Hang Zhao} and {A. Yuille} and {Yingwei Li}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/1358ad196c4e300612fb3b65a2f3578836941384}, } - T. Kocmi, R. Bawden, O. Bojar, A. Dvorkovich, C. Federmann, M. Fishel, T. Gowda, Y. Graham, R. Grundkiewicz, B. Haddow, R. Knowles, P. Koehn, C. Monz, M. Morishita, M. Nagata, T. Nakazawa, M. Novák, M. Popel, and M. Popović, “Findings of the 2022 Conference on Machine Translation (WMT22),” in Proceedings of the Seventh Conference on Machine Translation (WMT), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 1–45.
[BibTeX] [Abstract] [Link]
This paper presents the results of the General Machine Translation Task organised as part of the Conference on Machine Translation (WMT) 2022. In the general MT task, participants were asked to build machine translation systems for any of 11 language pairs, to be evaluated on test sets consisting of four different domains. We evaluate system outputs with human annotators using two different techniques: reference-based direct assessment and (DA) and a combination of DA and scalar quality metric (DA+SQM).
@inproceedings{kocmi-etal-2022-findings, title = "Findings of the 2022 Conference on Machine Translation ({WMT}22)", author = "Kocmi, Tom and Bawden, Rachel and Bojar, Ond{\v{r}}ej and Dvorkovich, Anton and Federmann, Christian and Fishel, Mark and Gowda, Thamme and Graham, Yvette and Grundkiewicz, Roman and Haddow, Barry and Knowles, Rebecca and Koehn, Philipp and Monz, Christof and Morishita, Makoto and Nagata, Masaaki and Nakazawa, Toshiaki and Nov{\'a}k, Michal and Popel, Martin and Popovi{\'c}, Maja", editor = {Koehn, Philipp and Barrault, Lo{\"\i}c and Bojar, Ond{\v{r}}ej and Bougares, Fethi and Chatterjee, Rajen and Costa-juss{\`a}, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Jimeno Yepes, Antonio and Kocmi, Tom and Martins, Andr{\'e} and Morishita, Makoto and Monz, Christof and Nagata, Masaaki and Nakazawa, Toshiaki and Negri, Matteo and N{\'e}v{\'e}ol, Aur{\'e}lie and Neves, Mariana and Popel, Martin and Turchi, Marco and Zampieri, Marcos}, booktitle = "Proceedings of the Seventh Conference on Machine Translation (WMT)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wmt-1.1", pages = "1--45", abstract = "This paper presents the results of the General Machine Translation Task organised as part of the Conference on Machine Translation (WMT) 2022. In the general MT task, participants were asked to build machine translation systems for any of 11 language pairs, to be evaluated on test sets consisting of four different domains. We evaluate system outputs with human annotators using two different techniques: reference-based direct assessment and (DA) and a combination of DA and scalar quality metric (DA+SQM).", } - Wenpin Hou, Mingyu Zhang, Yuelong Ji, X. Hong, Guoying Wang, Richard Xu, Liming Liang, S. Saria, and Hongkai Ji, “A prospective birth cohort study of maternal prenatal cigarette smoking assessed by self-report and biomarkers on childhood risk of overweight or obesity.,” in Precision Nutrition, 2022.
[BibTeX] [Link]@inproceedings{262383173, title = {A prospective birth cohort study of maternal prenatal cigarette smoking assessed by self-report and biomarkers on childhood risk of overweight or obesity.}, author = {{Wenpin Hou} and {Mingyu Zhang} and {Yuelong Ji} and {X. Hong} and {Guoying Wang} and {Richard Xu} and {Liming Liang} and {S. Saria} and {Hongkai Ji}}, year = 2022, month = {12}, booktitle = {Precision Nutrition}, url = {https://www.semanticscholar.org/paper/e898ce790bbb170c93ff44e139e83c3448b590ab}, } - Zhuowan Li, Xingrui Wang, Elias Stengel-Eskin, Adam Kortylewski, Wufei Ma, Benjamin Van Durme, Alan Yuille Johns Hopkins University, U. California, Max Planck Institute for Informatics, and U. Freiburg, “Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{254125164, title = {Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning}, author = {{Zhuowan Li} and {Xingrui Wang} and {Elias Stengel-Eskin} and {Adam Kortylewski} and {Wufei Ma} and {Benjamin Van Durme} and {Alan Yuille Johns Hopkins University} and {U. California} and {Max Planck Institute for Informatics} and {U. Freiburg}}, year = 2022, month = {12}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/cb6365a1aa3133318ce7fa2461b6d1d48cd8152e}, } - Wenpin Hou, Mingyu Zhang, Yuelong Ji, X. Hong, Guoying Wang, Richard Xu, L. Liang, S. Saria, and Hongkai Ji, “A prospective birth cohort study of maternal prenatal cigarette smoking assessed by self-report and biomarkers on childhood risk of overweight or obesity,” in Precision Nutrition, 2022.
[BibTeX] [Link]@inproceedings{254825419, title = {A prospective birth cohort study of maternal prenatal cigarette smoking assessed by self-report and biomarkers on childhood risk of overweight or obesity}, author = {{Wenpin Hou} and {Mingyu Zhang} and {Yuelong Ji} and {X. Hong} and {Guoying Wang} and {Richard Xu} and {L. Liang} and {S. Saria} and {Hongkai Ji}}, year = 2022, month = {12}, booktitle = {Precision Nutrition}, url = {https://www.semanticscholar.org/paper/2304edd25b8f08b99c6992c3de6434459742ccad}, } - Desh Raj, Daniel Povey, and S. Khudanpur, “GPU-accelerated Guided Source Separation for Meeting Transcription,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{254563914, title = {GPU-accelerated Guided Source Separation for Meeting Transcription}, author = {{Desh Raj} and {Daniel Povey} and {S. Khudanpur}}, year = 2022, month = {12}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/ad9199a35c5ce4738c39bb4af8abf7caee418365}, } - S. Sia, K. Jaidka, H. Ahuja, N. Chhaya, and K. Duh, “Offer a Different Perspective: Modeling the Belief Alignment of Arguments in Multi-party Debates,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 11939–11950. doi:10.18653/v1/2022.emnlp-main.818
[BibTeX] [Abstract] [Link]
In contexts where debate and deliberation are the norm, the participants are regularly presented with new information that conflicts with their original beliefs. When required to update their beliefs (belief alignment), they may choose arguments that align with their worldview (confirmation bias). We test this and competing hypotheses in a constraint-based modeling approach to predict the winning arguments in multi-party interactions in the Reddit Change My View and Intelligence Squared debates datasets. We adopt a hierarchical generative Variational Autoencoder as our model and impose structural constraints that reflect competing hypotheses about the nature of argumentation. Our findings suggest that in most settings, predictive models that anticipate winning arguments to be further from the initial argument of the opinion holder are more likely to succeed.
@inproceedings{sia-etal-2022-offer, title = "Offer a Different Perspective: Modeling the Belief Alignment of Arguments in Multi-party Debates", author = "Sia, Suzanna and Jaidka, Kokil and Ahuja, Hansin and Chhaya, Niyati and Duh, Kevin", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.818", doi = "10.18653/v1/2022.emnlp-main.818", pages = "11939--11950", abstract = "In contexts where debate and deliberation are the norm, the participants are regularly presented with new information that conflicts with their original beliefs. When required to update their beliefs (belief alignment), they may choose arguments that align with their worldview (confirmation bias). We test this and competing hypotheses in a constraint-based modeling approach to predict the winning arguments in multi-party interactions in the Reddit Change My View and Intelligence Squared debates datasets. We adopt a hierarchical generative Variational Autoencoder as our model and impose structural constraints that reflect competing hypotheses about the nature of argumentation. Our findings suggest that in most settings, predictive models that anticipate winning arguments to be further from the initial argument of the opinion holder are more likely to succeed.", } - M. Iglesias, A. Favaro, C. Motley, E. Oh, R. Stevens, A. Butala, L. Moro-Velázquez, and N. Dehak, “Cognitive and Acoustic Speech and Language Patterns Occurring in Different Neurodegenerative Disorders while Performing Neuropsychological Tests,” in IEEE Signal Processing in Medicine and Biology Symposium, 2022.
[BibTeX] [Link]@inproceedings{256033943, title = {Cognitive and Acoustic Speech and Language Patterns Occurring in Different Neurodegenerative Disorders while Performing Neuropsychological Tests}, author = {{M. Iglesias} and {A. Favaro} and {C. Motley} and {E. Oh} and {R. Stevens} and {A. Butala} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2022, month = {12}, booktitle = {IEEE Signal Processing in Medicine and Biology Symposium}, url = {https://www.semanticscholar.org/paper/ee067fbced756c332d18a34d6d4f59ab512f9013}, } - H. Kim, Y. Yu, L. Jiang, X. Lu, D. Khashabi, G. Kim, Y. Choi, and M. Sap, “ProsocialDialog: A Prosocial Backbone for Conversational Agents,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 4005–4029. doi:10.18653/v1/2022.emnlp-main.267
[BibTeX] [Abstract] [Link]
Most existing dialogue systems fail to respond properly to potentially unsafe user utterances by either ignoring or passively agreeing with them. To address this issue, we introduce ProsocialDialog, the first large-scale multi-turn dialogue dataset to teach conversational agents to respond to problematic content following social norms. Covering diverse unethical, problematic, biased, and toxic situations, ProsocialDialog contains responses that encourage prosocial behavior, grounded in commonsense social rules (i.e., rules-of-thumb, RoTs). Created via a human-AI collaborative framework, ProsocialDialog consists of 58K dialogues, with 331K utterances, 160K unique RoTs, and 497K dialogue safety labels accompanied by free-form rationales. With this dataset, we introduce a dialogue safety detection module, Canary, capable of generating RoTs given conversational context, and a socially-informed dialogue agent, Prost. Empirical results show that Prost generates more socially acceptable dialogues compared to other state-of-the-art language and dialogue models in both in-domain and out-of-domain settings. Additionally, Canary effectively guides conversational agents and off-the-shelf language models to generate significantly more prosocial responses. Our work highlights the promise and importance of creating and steering conversational AI to be socially responsible.
@inproceedings{kim-etal-2022-prosocialdialog, title = "{P}rosocial{D}ialog: A Prosocial Backbone for Conversational Agents", author = "Kim, Hyunwoo and Yu, Youngjae and Jiang, Liwei and Lu, Ximing and Khashabi, Daniel and Kim, Gunhee and Choi, Yejin and Sap, Maarten", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.267", doi = "10.18653/v1/2022.emnlp-main.267", pages = "4005--4029", abstract = "Most existing dialogue systems fail to respond properly to potentially unsafe user utterances by either ignoring or passively agreeing with them. To address this issue, we introduce ProsocialDialog, the first large-scale multi-turn dialogue dataset to teach conversational agents to respond to problematic content following social norms. Covering diverse unethical, problematic, biased, and toxic situations, ProsocialDialog contains responses that encourage prosocial behavior, grounded in commonsense social rules (i.e., rules-of-thumb, RoTs). Created via a human-AI collaborative framework, ProsocialDialog consists of 58K dialogues, with 331K utterances, 160K unique RoTs, and 497K dialogue safety labels accompanied by free-form rationales. With this dataset, we introduce a dialogue safety detection module, Canary, capable of generating RoTs given conversational context, and a socially-informed dialogue agent, Prost. Empirical results show that Prost generates more socially acceptable dialogues compared to other state-of-the-art language and dialogue models in both in-domain and out-of-domain settings. Additionally, Canary effectively guides conversational agents and off-the-shelf language models to generate significantly more prosocial responses. Our work highlights the promise and importance of creating and steering conversational AI to be socially responsible.", } - W. Gu, B. Zheng, Y. Chen, T. Chen, and B. Van Durme, “An Empirical Study on Finding Spans,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 3976–3983. doi:10.18653/v1/2022.emnlp-main.264
[BibTeX] [Abstract] [Link]
We present an empirical study on methods for span finding, the selection of consecutive tokens in text for some downstream tasks. We focus on approaches that can be employed in training end-to-end information extraction systems, and find there is no definitive solution without considering task properties, and provide our observations to help with future design choices: 1) a tagging approach often yields higher precision while span enumeration and boundary prediction provide higher recall; 2) span type information can benefit a boundary prediction approach; 3) additional contextualization does not help span finding in most cases.
@inproceedings{gu-etal-2022-empirical, title = "An Empirical Study on Finding Spans", author = "Gu, Weiwei and Zheng, Boyuan and Chen, Yunmo and Chen, Tongfei and Van Durme, Benjamin", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.264", doi = "10.18653/v1/2022.emnlp-main.264", pages = "3976--3983", abstract = "We present an empirical study on methods for span finding, the selection of consecutive tokens in text for some downstream tasks. We focus on approaches that can be employed in training end-to-end information extraction systems, and find there is no definitive solution without considering task properties, and provide our observations to help with future design choices: 1) a tagging approach often yields higher precision while span enumeration and boundary prediction provide higher recall; 2) span type information can benefit a boundary prediction approach; 3) additional contextualization does not help span finding in most cases.", } - Nathaniel Weir, Ryan Thomas, Randolph D’Amore, Kellie Hill, Benjamin Van Durme, and Harsh Jhamtani, “Ontologically Faithful Generation of Non-Player Character Dialogues,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254926993, title = {Ontologically Faithful Generation of Non-Player Character Dialogues}, author = {{Nathaniel Weir} and {Ryan Thomas} and {Randolph D'Amore} and {Kellie Hill} and {Benjamin Van Durme} and {Harsh Jhamtani}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/68f1b94bbc900d2b5c60192a7e9eea4b046dd18a}, } - M. Keymanesh, A. Benton, and M. Dredze, “What Makes Data-to-Text Generation Hard for Pretrained Language Models?,” in Proceedings of the 2nd Workshop on Natural Language Generation, Evaluation, and Metrics (GEM), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 539–554. doi:10.18653/v1/2022.gem-1.50
[BibTeX] [Abstract] [Link]
Expressing natural language descriptions of structured facts or relations {–} data-to-text generation (D2T) {–} increases the accessibility of structured knowledge repositories. Previous work shows that pre-trained language models (PLMs) perform remarkably well on this task after fine-tuning on a significant amount of task-specific training data. On the other hand, while auto-regressive PLMs can generalize from a few task examples, their efficacy at D2T is largely unexplored. Furthermore, we have an incomplete understanding of the limits of PLMs on D2T. In this work, we conduct an empirical study of both fine-tuned and auto-regressive PLMs on the DART multi-domain D2T dataset. We consider their performance as a function of the amount of task-specific data and how the data is incorporated into the models: zero and few-shot learning, and fine-tuning of model weights. In addition, we probe the limits of PLMs by measuring performance on subsets of the evaluation data: novel predicates and abstractive test examples. To improve the performance on these subsets, we investigate two techniques: providing predicate descriptions in the context and re-ranking generated candidates by information reflected in the source. Finally, we conduct a human evaluation of model errors and show that D2T generation tasks would benefit from datasets with more careful manual curation.
@inproceedings{keymanesh-etal-2022-makes, title = "What Makes Data-to-Text Generation Hard for Pretrained Language Models?", author = "Keymanesh, Moniba and Benton, Adrian and Dredze, Mark", editor = "Bosselut, Antoine and Chandu, Khyathi and Dhole, Kaustubh and Gangal, Varun and Gehrmann, Sebastian and Jernite, Yacine and Novikova, Jekaterina and Perez-Beltrachini, Laura", booktitle = "Proceedings of the 2nd Workshop on Natural Language Generation, Evaluation, and Metrics (GEM)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.gem-1.50", doi = "10.18653/v1/2022.gem-1.50", pages = "539--554", abstract = "Expressing natural language descriptions of structured facts or relations {--} data-to-text generation (D2T) {--} increases the accessibility of structured knowledge repositories. Previous work shows that pre-trained language models (PLMs) perform remarkably well on this task after fine-tuning on a significant amount of task-specific training data. On the other hand, while auto-regressive PLMs can generalize from a few task examples, their efficacy at D2T is largely unexplored. Furthermore, we have an incomplete understanding of the limits of PLMs on D2T. In this work, we conduct an empirical study of both fine-tuned and auto-regressive PLMs on the DART multi-domain D2T dataset. We consider their performance as a function of the amount of task-specific data and how the data is incorporated into the models: zero and few-shot learning, and fine-tuning of model weights. In addition, we probe the limits of PLMs by measuring performance on subsets of the evaluation data: novel predicates and abstractive test examples. To improve the performance on these subsets, we investigate two techniques: providing predicate descriptions in the context and re-ranking generated candidates by information reflected in the source. Finally, we conduct a human evaluation of model errors and show that D2T generation tasks would benefit from datasets with more careful manual curation.", } - K. Deb, X. Zhang, and K. Duh, “Post-Hoc Interpretation of Transformer Hyperparameters with Explainable Boosting Machines,” in Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 51–61. doi:10.18653/v1/2022.blackboxnlp-1.5
[BibTeX] [Abstract] [Link]
Hyperparameter tuning is important for achieving high accuracy in deep learning models, yet little interpretability work has focused on hyperparameters. We propose to use the Explainable Boosting Machine (EBM), a glassbox method, as a post-hoc analysis tool for understanding how hyperparameters influence model accuracy. We present a case study on Transformer models in machine translation to illustrate the kinds of insights that may be gleaned, and perform extensive analysis to test the robustness of EBM under different data conditions.
@inproceedings{deb-etal-2022-post, title = "Post-Hoc Interpretation of Transformer Hyperparameters with Explainable Boosting Machines", author = "Deb, Kiron and Zhang, Xuan and Duh, Kevin", editor = "Bastings, Jasmijn and Belinkov, Yonatan and Elazar, Yanai and Hupkes, Dieuwke and Saphra, Naomi and Wiegreffe, Sarah", booktitle = "Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.blackboxnlp-1.5", doi = "10.18653/v1/2022.blackboxnlp-1.5", pages = "51--61", abstract = "Hyperparameter tuning is important for achieving high accuracy in deep learning models, yet little interpretability work has focused on hyperparameters. We propose to use the Explainable Boosting Machine (EBM), a glassbox method, as a post-hoc analysis tool for understanding how hyperparameters influence model accuracy. We present a case study on Transformer models in machine translation to illustrate the kinds of insights that may be gleaned, and perform extensive analysis to test the robustness of EBM under different data conditions.", } - Kangda Wei, Dawn J Lawrie, Benjamin Van Durme, Yunmo Chen, and Orion Weller, “When Do Decompositions Help for Machine Reading?,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254877586, title = {When Do Decompositions Help for Machine Reading?}, author = {{Kangda Wei} and {Dawn J Lawrie} and {Benjamin Van Durme} and {Yunmo Chen} and {Orion Weller}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/624ea7bdaf7e8e3f7bd76f72aa665b562f0dd70a}, } - Zhuowan Li, Cihang Xie, Benjamin Van Durme, Alan Yuille Johns Hopkins University, U. California, and Santa Cruz, “Localization vs. Semantics: How Can Language Benefit Visual Representation Learning?,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254125621, title = {Localization vs. Semantics: How Can Language Benefit Visual Representation Learning?}, author = {{Zhuowan Li} and {Cihang Xie} and {Benjamin Van Durme} and {Alan Yuille Johns Hopkins University} and {U. California} and {Santa Cruz}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/970a8ed9de244b080aa69dbf5996a37057909ca6}, } - R. Wicks and M. Post, “Does Sentence Segmentation Matter for Machine Translation?,” in Proceedings of the Seventh Conference on Machine Translation (WMT), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 843–854.
[BibTeX] [Abstract] [Link]
For the most part, NLP applications operate at the sentence level. Since sentences occur most naturally in documents, they must be extracted and segmented via the use of a segmenter, of which there are a handful of options. There has been some work evaluating the performance of segmenters on intrinsic metrics, that look at their ability to recover human-segmented sentence boundaries, but there has been no work looking at the effect of segmenters on downstream tasks. We ask the question, {“}does segmentation matter?{”} and attempt to answer it on the task of machine translation. We consider two settings: the application of segmenters to a black-box system whose training segmentation is mostly unknown, as well as the variation in performance when segmenters are applied to the training process, too. We find that the choice of segmenter largely does not matter, so long as its behavior is not one of extreme under- or over-segmentation. For such settings, we provide some qualitative analysis examining their harms, and point the way towards document-level processing.
@inproceedings{wicks-post-2022-sentence, title = "Does Sentence Segmentation Matter for Machine Translation?", author = "Wicks, Rachel and Post, Matt", editor = {Koehn, Philipp and Barrault, Lo{\"\i}c and Bojar, Ond{\v{r}}ej and Bougares, Fethi and Chatterjee, Rajen and Costa-juss{\`a}, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Jimeno Yepes, Antonio and Kocmi, Tom and Martins, Andr{\'e} and Morishita, Makoto and Monz, Christof and Nagata, Masaaki and Nakazawa, Toshiaki and Negri, Matteo and N{\'e}v{\'e}ol, Aur{\'e}lie and Neves, Mariana and Popel, Martin and Turchi, Marco and Zampieri, Marcos}, booktitle = "Proceedings of the Seventh Conference on Machine Translation (WMT)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wmt-1.78", pages = "843--854", abstract = "For the most part, NLP applications operate at the sentence level. Since sentences occur most naturally in documents, they must be extracted and segmented via the use of a segmenter, of which there are a handful of options. There has been some work evaluating the performance of segmenters on intrinsic metrics, that look at their ability to recover human-segmented sentence boundaries, but there has been no work looking at the effect of segmenters on downstream tasks. We ask the question, {``}does segmentation matter?{''} and attempt to answer it on the task of machine translation. We consider two settings: the application of segmenters to a black-box system whose training segmentation is mostly unknown, as well as the variation in performance when segmenters are applied to the training process, too. We find that the choice of segmenter largely does not matter, so long as its behavior is not one of extreme under- or over-segmentation. For such settings, we provide some qualitative analysis examining their harms, and point the way towards document-level processing.", } - C. Chen, L. Moro-Velázquez, A. Ožbolt, A. Butala, A. Pantelyat, and N. Dehak, “Phonatory Analysis on Parkinson’s Disease Patients Attending Singing and Discussion Therapy (Parkinsonics) using Signal Processing Techniques,” in IEEE Signal Processing in Medicine and Biology Symposium, 2022.
[BibTeX] [Link]@inproceedings{256034037, title = {Phonatory Analysis on Parkinson's Disease Patients Attending Singing and Discussion Therapy (Parkinsonics) using Signal Processing Techniques}, author = {{C. Chen} and {L. Moro-Velázquez} and {A. Ožbolt} and {A. Butala} and {A. Pantelyat} and {N. Dehak}}, year = 2022, month = {12}, booktitle = {IEEE Signal Processing in Medicine and Biology Symposium}, url = {https://www.semanticscholar.org/paper/513937e2300445136193356fb6fdae3753d09770}, } - O. Ogundepo, X. Zhang, S. Sun, K. Duh, and J. Lin, “AfriCLIRMatrix: Enabling Cross-Lingual Information Retrieval for African Languages,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 8721–8728. doi:10.18653/v1/2022.emnlp-main.597
[BibTeX] [Abstract] [Link]
Language diversity in NLP is critical in enabling the development of tools for a wide range of users. However, there are limited resources for building such tools for many languages, particularly those spoken in Africa.For search, most existing datasets feature few or no African languages, directly impacting researchers{‘} ability to build and improve information access capabilities in those languages. Motivated by this, we created AfriCLIRMatrix, a test collection for cross-lingual information retrieval research in 15 diverse African languages. In total, our dataset contains 6 million queries in English and 23 million relevance judgments automatically mined from Wikipedia inter-language links, covering many more African languages than any existing information retrieval test collection. In addition, we release BM25, dense retrieval, and sparse{–}dense hybrid baselines to provide a starting point for the development of future systems. We hope that these efforts can spur additional work in search for African languages.AfriCLIRMatrix can be downloaded at \url{https://github.com/castorini/africlirmatrix}.
@inproceedings{ogundepo-etal-2022-africlirmatrix, title = "{A}fri{CLIRM}atrix: Enabling Cross-Lingual Information Retrieval for {A}frican Languages", author = "Ogundepo, Odunayo and Zhang, Xinyu and Sun, Shuo and Duh, Kevin and Lin, Jimmy", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.597", doi = "10.18653/v1/2022.emnlp-main.597", pages = "8721--8728", abstract = "Language diversity in NLP is critical in enabling the development of tools for a wide range of users. However, there are limited resources for building such tools for many languages, particularly those spoken in Africa.For search, most existing datasets feature few or no African languages, directly impacting researchers{'} ability to build and improve information access capabilities in those languages. Motivated by this, we created AfriCLIRMatrix, a test collection for cross-lingual information retrieval research in 15 diverse African languages. In total, our dataset contains 6 million queries in English and 23 million relevance judgments automatically mined from Wikipedia inter-language links, covering many more African languages than any existing information retrieval test collection. In addition, we release BM25, dense retrieval, and sparse{--}dense hybrid baselines to provide a starting point for the development of future systems. We hope that these efforts can spur additional work in search for African languages.AfriCLIRMatrix can be downloaded at \url{https://github.com/castorini/africlirmatrix}.", } - E. Stengel-Eskin and B. Van Durme, “The Curious Case of Control,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 11065–11076. doi:10.18653/v1/2022.emnlp-main.760
[BibTeX] [Abstract] [Link]
Children acquiring English make systematic errors on subject control sentences even after they have reached near-adult competence (Chomsky, 1969), possibly due to heuristics based on semantic roles (Maratsos, 1974).Given the advanced fluency of large generative language models, we ask whether model outputs are consistent with these heuristics, and to what degree different models are consistent with each other. We find that models can be categorized by behavior into three separate groups, with broad differences between the groups. The outputs of models in the largest group are consistent with positional heuristics that succeed on subject control but fail on object control. This result is surprising, given that object control is orders of magnitude more frequent in the text data used to train such models. We examine to what degree the models are sensitive to prompting with agent-patient information, finding that raising the salience of agent and patient relations results in significant changes in the outputs of most models. Based on this observation, we leverage an existing dataset of semantic proto-role annotations (White et al. 2020) to explore the connections between control and labeling event participants with properties typically associated with agents and patients.
@inproceedings{stengel-eskin-van-durme-2022-curious, title = "The Curious Case of Control", author = "Stengel-Eskin, Elias and Van Durme, Benjamin", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.760", doi = "10.18653/v1/2022.emnlp-main.760", pages = "11065--11076", abstract = "Children acquiring English make systematic errors on subject control sentences even after they have reached near-adult competence (Chomsky, 1969), possibly due to heuristics based on semantic roles (Maratsos, 1974).Given the advanced fluency of large generative language models, we ask whether model outputs are consistent with these heuristics, and to what degree different models are consistent with each other. We find that models can be categorized by behavior into three separate groups, with broad differences between the groups. The outputs of models in the largest group are consistent with positional heuristics that succeed on subject control but fail on object control. This result is surprising, given that object control is orders of magnitude more frequent in the text data used to train such models. We examine to what degree the models are sensitive to prompting with agent-patient information, finding that raising the salience of agent and patient relations results in significant changes in the outputs of most models. Based on this observation, we leverage an existing dataset of semantic proto-role annotations (White et al. 2020) to explore the connections between control and labeling event participants with properties typically associated with agents and patients.", } - Z. Jiang, A. Liu, and B. Van Durme, “Calibrating Zero-shot Cross-lingual (Un-)structured Predictions,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 2648–2674. doi:10.18653/v1/2022.emnlp-main.170
[BibTeX] [Abstract] [Link]
We investigate model calibration in the setting of zero-shot cross-lingual transfer with large-scale pre-trained language models. The level of model calibration is an important metric for evaluating the trustworthiness of predictive models. There exists an essential need for model calibration when natural language models are deployed in critical tasks. We study different post-training calibration methods in structured and unstructured prediction tasks. We find that models trained with data from the source language become less calibrated when applied to the target language and that calibration errors increase with intrinsic task difficulty and relative sparsity of training data. Moreover, we observe a potential connection between the level of calibration error and an earlier proposed measure of the distance from English to other languages. Finally, our comparison demonstrates that among other methods Temperature Scaling (TS) generalizes well to distant languages, but TS fails to calibrate more complex confidence estimation in structured predictions compared to more expressive alternatives like Gaussian Process Calibration.
@inproceedings{jiang-etal-2022-calibrating, title = "Calibrating Zero-shot Cross-lingual (Un-)structured Predictions", author = "Jiang, Zhengping and Liu, Anqi and Van Durme, Benjamin", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.170", doi = "10.18653/v1/2022.emnlp-main.170", pages = "2648--2674", abstract = "We investigate model calibration in the setting of zero-shot cross-lingual transfer with large-scale pre-trained language models. The level of model calibration is an important metric for evaluating the trustworthiness of predictive models. There exists an essential need for model calibration when natural language models are deployed in critical tasks. We study different post-training calibration methods in structured and unstructured prediction tasks. We find that models trained with data from the source language become less calibrated when applied to the target language and that calibration errors increase with intrinsic task difficulty and relative sparsity of training data. Moreover, we observe a potential connection between the level of calibration error and an earlier proposed measure of the distance from English to other languages. Finally, our comparison demonstrates that among other methods Temperature Scaling (TS) generalizes well to distant languages, but TS fails to calibrate more complex confidence estimation in structured predictions compared to more expressive alternatives like Gaussian Process Calibration.", } - Nithin Gopalakrishnan Nair, W. G. C. Bandara, and Vishal M. Patel, “Unite and Conquer: Cross Dataset Multimodal Synthesis using Diffusion Models,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254125357, title = {Unite and Conquer: Cross Dataset Multimodal Synthesis using Diffusion Models}, author = {{Nithin Gopalakrishnan Nair} and {W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/24e6c62fd28da4ecf748620e1f25eae7337bad40}, } - Junyang Wu, Xianhang Li, Chen Wei, Huiyu Wang, A. Yuille, Yuyin Zhou, and Cihang Xie, “Unleashing the Power of Visual Prompting At the Pixel Level,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254877370, title = {Unleashing the Power of Visual Prompting At the Pixel Level}, author = {{Junyang Wu} and {Xianhang Li} and {Chen Wei} and {Huiyu Wang} and {A. Yuille} and {Yuyin Zhou} and {Cihang Xie}}, year = 2022, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7786825fd653b398c3975c3ff876459307d871f4}, } - Y. Wang, S. Mishra, P. Alipoormolabashi, Y. Kordi, A. Mirzaei, A. Naik, A. Ashok, A. S. Dhanasekaran, A. Arunkumar, D. Stap, E. Pathak, G. Karamanolakis, H. Lai, I. Purohit, I. Mondal, J. Anderson, K. Kuznia, K. Doshi, K. K. Pal, M. Patel, M. Moradshahi, M. Parmar, M. Purohit, N. Varshney, P. R. Kaza, P. Verma, R. S. Puri, R. Karia, S. Doshi, S. K. Sampat, S. Mishra, S. Reddy A, S. Patro, T. Dixit, and X. Shen, “Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 5085–5109. doi:10.18653/v1/2022.emnlp-main.340
[BibTeX] [Abstract] [Link]
How well can NLP models generalize to a variety of unseen tasks when provided with task instructions? To address this question, we first introduce Super-NaturalInstructions, a benchmark of 1,616 diverse NLP tasks and their expert-written instructions. Our collection covers 76 distinct task types, including but not limited to classification, extraction, infilling, sequence tagging, text rewriting, and text composition. This large and diverse collection of tasks enables rigorous benchmarking of cross-task generalization under instructions{–-}training models to follow instructions on a subset of tasks and evaluating them on the remaining unseen ones. Furthermore, we build Tk-Instruct, a transformer model trained to follow a variety of in-context instructions (plain language task definitions or k-shot examples). Our experiments show that Tk-Instruct outperforms existing instruction-following models such as InstructGPT by over 9{\%} on our benchmark despite being an order of magnitude smaller. We further analyze generalization as a function of various scaling parameters, such as the number of observed tasks, the number of instances per task, and model sizes. We hope our dataset and model facilitate future progress towards more general-purpose NLP models.
@inproceedings{wang-etal-2022-super, title = "Super-{N}atural{I}nstructions: Generalization via Declarative Instructions on 1600+ {NLP} Tasks", author = "Wang, Yizhong and Mishra, Swaroop and Alipoormolabashi, Pegah and Kordi, Yeganeh and Mirzaei, Amirreza and Naik, Atharva and Ashok, Arjun and Dhanasekaran, Arut Selvan and Arunkumar, Anjana and Stap, David and Pathak, Eshaan and Karamanolakis, Giannis and Lai, Haizhi and Purohit, Ishan and Mondal, Ishani and Anderson, Jacob and Kuznia, Kirby and Doshi, Krima and Pal, Kuntal Kumar and Patel, Maitreya and Moradshahi, Mehrad and Parmar, Mihir and Purohit, Mirali and Varshney, Neeraj and Kaza, Phani Rohitha and Verma, Pulkit and Puri, Ravsehaj Singh and Karia, Rushang and Doshi, Savan and Sampat, Shailaja Keyur and Mishra, Siddhartha and Reddy A, Sujan and Patro, Sumanta and Dixit, Tanay and Shen, Xudong", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.340", doi = "10.18653/v1/2022.emnlp-main.340", pages = "5085--5109", abstract = "How well can NLP models generalize to a variety of unseen tasks when provided with task instructions? To address this question, we first introduce Super-NaturalInstructions, a benchmark of 1,616 diverse NLP tasks and their expert-written instructions. Our collection covers 76 distinct task types, including but not limited to classification, extraction, infilling, sequence tagging, text rewriting, and text composition. This large and diverse collection of tasks enables rigorous benchmarking of cross-task generalization under instructions{---}training models to follow instructions on a subset of tasks and evaluating them on the remaining unseen ones. Furthermore, we build Tk-Instruct, a transformer model trained to follow a variety of in-context instructions (plain language task definitions or k-shot examples). Our experiments show that Tk-Instruct outperforms existing instruction-following models such as InstructGPT by over 9{\%} on our benchmark despite being an order of magnitude smaller. We further analyze generalization as a function of various scaling parameters, such as the number of observed tasks, the number of instances per task, and model sizes. We hope our dataset and model facilitate future progress towards more general-purpose NLP models.", } - E. Schumacher, J. Mayfield, and M. Dredze, “Zero-shot Cross-Language Transfer of Monolingual Entity Linking Models,” in Proceedings of the The 2nd Workshop on Multi-lingual Representation Learning (MRL), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 38–51. doi:10.18653/v1/2022.mrl-1.4
[BibTeX] [Abstract] [Link]
Most entity linking systems, whether mono or multilingual, link mentions to a single English knowledge base. Few have considered linking non-English text to a non-English KB, and therefore, transferring an English entity linking model to both a new document and KB language. We consider the task of zero-shot cross-language transfer of entity linking systems to a new language and KB. We find that a system trained with multilingual representations does reasonably well, and propose improvements to system training that lead to improved recall in most datasets, often matching the in-language performance. We further conduct a detailed evaluation to elucidate the challenges of this setting.
@inproceedings{schumacher-etal-2022-zero, title = "Zero-shot Cross-Language Transfer of Monolingual Entity Linking Models", author = "Schumacher, Elliot and Mayfield, James and Dredze, Mark", editor = {Ataman, Duygu and Gonen, Hila and Ruder, Sebastian and Firat, Orhan and G{\"u}l Sahin, G{\"o}zde and Mirzakhalov, Jamshidbek}, booktitle = "Proceedings of the The 2nd Workshop on Multi-lingual Representation Learning (MRL)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.mrl-1.4", doi = "10.18653/v1/2022.mrl-1.4", pages = "38--51", abstract = "Most entity linking systems, whether mono or multilingual, link mentions to a single English knowledge base. Few have considered linking non-English text to a non-English KB, and therefore, transferring an English entity linking model to both a new document and KB language. We consider the task of zero-shot cross-language transfer of entity linking systems to a new language and KB. We find that a system trained with multilingual representations does reasonably well, and propose improvements to system training that lead to improved recall in most datasets, often matching the in-language performance. We further conduct a detailed evaluation to elucidate the challenges of this setting.", } - E. Rippeth and M. Post, “Additive Interventions Yield Robust Multi-Domain Machine Translation Models,” in Proceedings of the Seventh Conference on Machine Translation (WMT), Abu Dhabi, United Arab Emirates (Hybrid), 2022, p. 220–232.
[BibTeX] [Abstract] [Link]
Additive interventions are a recently-proposed mechanism for controlling target-side attributes in neural machine translation by modulating the encoder{‘}s representation of a source sequence as opposed to manipulating the raw source sequence as seen in most previous tag-based approaches. In this work we examine the role of additive interventions in a large-scale multi-domain machine translation setting and compare its performance in various inference scenarios. We find that while the performance difference is small between intervention-based systems and tag-based systems when the domain label matches the test domain, intervention-based systems are robust to label error, making them an attractive choice under label uncertainty. Further, we find that the superiority of single-domain fine-tuning comes under question when training data is scaled, contradicting previous findings.
@inproceedings{rippeth-post-2022-additive, title = "Additive Interventions Yield Robust Multi-Domain Machine Translation Models", author = "Rippeth, Elijah and Post, Matt", editor = {Koehn, Philipp and Barrault, Lo{\"\i}c and Bojar, Ond{\v{r}}ej and Bougares, Fethi and Chatterjee, Rajen and Costa-juss{\`a}, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Jimeno Yepes, Antonio and Kocmi, Tom and Martins, Andr{\'e} and Morishita, Makoto and Monz, Christof and Nagata, Masaaki and Nakazawa, Toshiaki and Negri, Matteo and N{\'e}v{\'e}ol, Aur{\'e}lie and Neves, Mariana and Popel, Martin and Turchi, Marco and Zampieri, Marcos}, booktitle = "Proceedings of the Seventh Conference on Machine Translation (WMT)", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wmt-1.14", pages = "220--232", abstract = "Additive interventions are a recently-proposed mechanism for controlling target-side attributes in neural machine translation by modulating the encoder{'}s representation of a source sequence as opposed to manipulating the raw source sequence as seen in most previous tag-based approaches. In this work we examine the role of additive interventions in a large-scale multi-domain machine translation setting and compare its performance in various inference scenarios. We find that while the performance difference is small between intervention-based systems and tag-based systems when the domain label matches the test domain, intervention-based systems are robust to label error, making them an attractive choice under label uncertainty. Further, we find that the superiority of single-domain fine-tuning comes under question when training data is scaled, contradicting previous findings.", } - H. Xu, P. Koehn, and K. Murray, “The Importance of Being Parameters: An Intra-Distillation Method for Serious Gains,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 170–183. doi:10.18653/v1/2022.emnlp-main.13
[BibTeX] [Abstract] [Link]
Recent model pruning methods have demonstrated the ability to remove redundant parameters without sacrificing model performance. Common methods remove redundant parameters according to the parameter sensitivity, a gradient-based measure reflecting the contribution of the parameters. In this paper, however, we argue that redundant parameters can be trained to make beneficial contributions. We first highlight the large sensitivity (contribution) gap among high-sensitivity and low-sensitivity parameters and show that the model generalization performance can be significantly improved after balancing the contribution of all parameters. Our goal is to balance the sensitivity of all parameters and encourage all of them to contribute equally. We propose a general task-agnostic method, namely intra-distillation, appended to the regular training loss to balance parameter sensitivity. Moreover, we also design a novel adaptive learning method to control the strength of intra-distillation loss for faster convergence. Our experiments show the strong effectiveness of our methods on machine translation, natural language understanding, and zero-shot cross-lingual transfer across up to 48 languages, e.g., a gain of 3.54 BLEU on average across 8 language pairs from the IWSLT{‘}14 dataset.
@inproceedings{xu-etal-2022-importance, title = "The Importance of Being Parameters: An Intra-Distillation Method for Serious Gains", author = "Xu, Haoran and Koehn, Philipp and Murray, Kenton", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.13", doi = "10.18653/v1/2022.emnlp-main.13", pages = "170--183", abstract = "Recent model pruning methods have demonstrated the ability to remove redundant parameters without sacrificing model performance. Common methods remove redundant parameters according to the parameter sensitivity, a gradient-based measure reflecting the contribution of the parameters. In this paper, however, we argue that redundant parameters can be trained to make beneficial contributions. We first highlight the large sensitivity (contribution) gap among high-sensitivity and low-sensitivity parameters and show that the model generalization performance can be significantly improved after balancing the contribution of all parameters. Our goal is to balance the sensitivity of all parameters and encourage all of them to contribute equally. We propose a general task-agnostic method, namely intra-distillation, appended to the regular training loss to balance parameter sensitivity. Moreover, we also design a novel adaptive learning method to control the strength of intra-distillation loss for faster convergence. Our experiments show the strong effectiveness of our methods on machine translation, natural language understanding, and zero-shot cross-lingual transfer across up to 48 languages, e.g., a gain of 3.54 BLEU on average across 8 language pairs from the IWSLT{'}14 dataset.", } - K. Marchisio, N. Verma, K. Duh, and P. Koehn, “IsoVec: Controlling the Relative Isomorphism of Word Embedding Spaces,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 6019–6033. doi:10.18653/v1/2022.emnlp-main.404
[BibTeX] [Abstract] [Link]
The ability to extract high-quality translation dictionaries from monolingual word embedding spaces depends critically on the geometric similarity of the spaces{–-}their degree of {“}isomorphism.{”} We address the root-cause of faulty cross-lingual mapping: that word embedding training resulted in the underlying spaces being non-isomorphic. We incorporate global measures of isomorphism directly into the skipgram loss function, successfully increasing the relative isomorphism of trained word embedding spaces and improving their ability to be mapped to a shared cross-lingual space. The result is improved bilingual lexicon induction in general data conditions, under domain mismatch, and with training algorithm dissimilarities. We release IsoVec at \url{https://github.com/kellymarchisio/isovec}.
@inproceedings{marchisio-etal-2022-isovec, title = "{I}so{V}ec: Controlling the Relative Isomorphism of Word Embedding Spaces", author = "Marchisio, Kelly and Verma, Neha and Duh, Kevin and Koehn, Philipp", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.404", doi = "10.18653/v1/2022.emnlp-main.404", pages = "6019--6033", abstract = "The ability to extract high-quality translation dictionaries from monolingual word embedding spaces depends critically on the geometric similarity of the spaces{---}their degree of {``}isomorphism.{''} We address the root-cause of faulty cross-lingual mapping: that word embedding training resulted in the underlying spaces being non-isomorphic. We incorporate global measures of isomorphism directly into the skipgram loss function, successfully increasing the relative isomorphism of trained word embedding spaces and improving their ability to be mapped to a shared cross-lingual space. The result is improved bilingual lexicon induction in general data conditions, under domain mismatch, and with training algorithm dissimilarities. We release IsoVec at \url{https://github.com/kellymarchisio/isovec}.", } - E. Stengel-Eskin, E. A. Platanios, A. Pauls, S. Thomson, H. Fang, B. Van Durme, J. Eisner, and Y. Su, “When More Data Hurts: A Troubling Quirk in Developing Broad-Coverage Natural Language Understanding Systems,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 11473–11487. doi:10.18653/v1/2022.emnlp-main.789
[BibTeX] [Abstract] [Link]
In natural language understanding (NLU) production systems, users{‘} evolving needs necessitate the addition of new features over time, indexed by new symbols added to the meaning representation space. This requires additional training data and results in ever-growing datasets. We present the first systematic investigation into this incremental symbol learning scenario. Our analysis reveals a troubling quirk in building broad-coverage NLU systems: as the training dataset grows, performance on a small set of new symbols often decreases. We show that this trend holds for multiple mainstream models on two common NLU tasks: intent recognition and semantic parsing. Rejecting class imbalance as the sole culprit, we reveal that the trend is closely associated with an effect we call source signal dilution, where strong lexical cues for the new symbol become diluted as the training dataset grows. Selectively dropping training examples to prevent dilution often reverses the trend, showing the over-reliance of mainstream neural NLU models on simple lexical cues.
@inproceedings{stengel-eskin-etal-2022-data, title = "When More Data Hurts: A Troubling Quirk in Developing Broad-Coverage Natural Language Understanding Systems", author = "Stengel-Eskin, Elias and Platanios, Emmanouil Antonios and Pauls, Adam and Thomson, Sam and Fang, Hao and Van Durme, Benjamin and Eisner, Jason and Su, Yu", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.789", doi = "10.18653/v1/2022.emnlp-main.789", pages = "11473--11487", abstract = "In natural language understanding (NLU) production systems, users{'} evolving needs necessitate the addition of new features over time, indexed by new symbols added to the meaning representation space. This requires additional training data and results in ever-growing datasets. We present the first systematic investigation into this incremental symbol learning scenario. Our analysis reveals a troubling quirk in building broad-coverage NLU systems: as the training dataset grows, performance on a small set of new symbols often decreases. We show that this trend holds for multiple mainstream models on two common NLU tasks: intent recognition and semantic parsing. Rejecting class imbalance as the sole culprit, we reveal that the trend is closely associated with an effect we call source signal dilution, where strong lexical cues for the new symbol become diluted as the training dataset grows. Selectively dropping training examples to prevent dilution often reverses the trend, showing the over-reliance of mainstream neural NLU models on simple lexical cues.", } - David Mueller, Nicholas Andrews, and Mark Dredze, “Do Text-to-Text Multi-Task Learners Suffer from Task Conflict?,” in Conference on Empirical Methods in Natural Language Processing, 2022.
[BibTeX] [Link]@inproceedings{254591386, title = {Do Text-to-Text Multi-Task Learners Suffer from Task Conflict?}, author = {{David Mueller} and {Nicholas Andrews} and {Mark Dredze}}, year = 2022, month = {12}, booktitle = {Conference on Empirical Methods in Natural Language Processing}, url = {https://www.semanticscholar.org/paper/2843661ee0d5fa159165beba50c345566cc44c57}, } - Y. Feng, F. Li, and P. Koehn, “Toward the Limitation of Code-Switching in Cross-Lingual Transfer,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 5966–5971. doi:10.18653/v1/2022.emnlp-main.400
[BibTeX] [Abstract] [Link]
Multilingual pretrained models have shown strong cross-lingual transfer ability. Some works used code-switching sentences, which consist of tokens from multiple languages, to enhance the cross-lingual representation further, and have shown success in many zero-shot cross-lingual tasks. However, code-switched tokens are likely to cause grammatical incoherence in newly substituted sentences, and negatively affect the performance on token-sensitive tasks, such as Part-of-Speech (POS) tagging and Named-Entity-Recognition (NER). This paper mitigates the limitation of the code-switching method by not only making the token replacement but considering the similarity between the context and the switched tokens so that the newly substituted sentences are grammatically consistent during both training and inference. We conduct experiments on cross-lingual POS and NER over 30+ languages, and demonstrate the effectiveness of our method by outperforming the mBERT by 0.95 and original code-switching method by 1.67 on F1 scores.
@inproceedings{feng-etal-2022-toward, title = "Toward the Limitation of Code-Switching in Cross-Lingual Transfer", author = "Feng, Yukun and Li, Feng and Koehn, Philipp", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.400", doi = "10.18653/v1/2022.emnlp-main.400", pages = "5966--5971", abstract = "Multilingual pretrained models have shown strong cross-lingual transfer ability. Some works used code-switching sentences, which consist of tokens from multiple languages, to enhance the cross-lingual representation further, and have shown success in many zero-shot cross-lingual tasks. However, code-switched tokens are likely to cause grammatical incoherence in newly substituted sentences, and negatively affect the performance on token-sensitive tasks, such as Part-of-Speech (POS) tagging and Named-Entity-Recognition (NER). This paper mitigates the limitation of the code-switching method by not only making the token replacement but considering the similarity between the context and the switched tokens so that the newly substituted sentences are grammatically consistent during both training and inference. We conduct experiments on cross-lingual POS and NER over 30+ languages, and demonstrate the effectiveness of our method by outperforming the mBERT by 0.95 and original code-switching method by 1.67 on F1 scores.", } - Orion Weller, Aleem Khan, Nathaniel Weir, Dawn J Lawrie, and Benjamin Van Durme, “Defending Against Misinformation Attacks in Open-Domain Question Answering.” 2022.
[BibTeX] [Link]@inproceedings{258685264, title = {Defending Against Misinformation Attacks in Open-Domain Question Answering}, author = {{Orion Weller} and {Aleem Khan} and {Nathaniel Weir} and {Dawn J Lawrie} and {Benjamin Van Durme}}, year = 2022, month = {12}, booktitle = {}, url = {https://www.semanticscholar.org/paper/55dca1a431f3de1fc3abceb6d5ff1d424936dd6c}, } - K. Marchisio, A. Saad-Eldin, K. Duh, C. Priebe, and P. Koehn, “Bilingual Lexicon Induction for Low-Resource Languages using Graph Matching via Optimal Transport,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 2545–2561. doi:10.18653/v1/2022.emnlp-main.164
[BibTeX] [Abstract] [Link]
Bilingual lexicons form a critical component of various natural language processing applications, including unsupervised and semisupervised machine translation and crosslingual information retrieval. In this work, we improve bilingual lexicon induction performance across 40 language pairs with a graph-matching method based on optimal transport. The method is especially strong with low amounts of supervision.
@inproceedings{marchisio-etal-2022-bilingual, title = "Bilingual Lexicon Induction for Low-Resource Languages using Graph Matching via Optimal Transport", author = "Marchisio, Kelly and Saad-Eldin, Ali and Duh, Kevin and Priebe, Carey and Koehn, Philipp", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.164", doi = "10.18653/v1/2022.emnlp-main.164", pages = "2545--2561", abstract = "Bilingual lexicons form a critical component of various natural language processing applications, including unsupervised and semisupervised machine translation and crosslingual information retrieval. In this work, we improve bilingual lexicon induction performance across 40 language pairs with a graph-matching method based on optimal transport. The method is especially strong with low amounts of supervision.", } - A. Svete, B. Dayan, R. Cotterell, T. Vieira, and J. Eisner, “Acyclic Weighted Finite-State Automata with Failure Transitions,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, 2022, p. 8289–8305.
[BibTeX] [Link]@InProceedings{svete-et-al-2022, aclid = "2022.emnlp-main.567", author = "Anej Svete and Benjamin Dayan and Ryan Cotterell and Tim Vieira and Jason Eisner", title = "Acyclic Weighted Finite-State Automata with Failure Transitions", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", pages = "8289--8305", year = "2022", month = dec, address = "Abu Dhabi", URL = "http://cs.jhu.edu/~jason/papers/#svete-et-al-2022", } - E. Stengel-Eskin, E. A. Platanios, A. Pauls, S. Thomson, H. Fang, B. V. Durme, J. Eisner, and Y. Su, “When More Data Hurts: A Troubling Quirk in Developing Broad-Coverage Natural Language Understanding Systems,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, 2022, p. 11473–11487.
[BibTeX] [Link]@InProceedings{stengeleskin-et-al-2022, aclid = "2022.emnlp-main.789", author = "Elias Stengel-Eskin and Emmanouil Antonios Platanios and Adam Pauls and Sam Thomson and Hao Fang and Benjamin Van Durme and Jason Eisner and Yu Su", title = "When More Data Hurts: {A} Troubling Quirk in Developing Broad-Coverage Natural Language Understanding Systems", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", pages = "11473--11487", year = "2022", month = dec, address = "Abu Dhabi", URL = "http://cs.jhu.edu/~jason/papers/#stengeleskin-et-al-2022", } - Jiatong Shi, Chan-Jan Hsu, Ho-Lam Chung, Dongji Gao, Leibny Paola García-Perera, Shinji Watanabe, Ann Lee, and Hung-yi Lee, “Bridging Speech and Textual Pre-Trained Models With Unsupervised ASR,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{253383773, title = {Bridging Speech and Textual Pre-Trained Models With Unsupervised ASR}, author = {{Jiatong Shi} and {Chan-Jan Hsu} and {Ho-Lam Chung} and {Dongji Gao} and {Leibny Paola García-Perera} and {Shinji Watanabe} and {Ann Lee} and {Hung-yi Lee}}, year = 2022, month = {11}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/92302ab168429c7c3a8f699b35ba8302916c6e7c}, } - Vikas Raunak, Matt Post, and Arul Menezes, “Operationalizing Specifications, In Addition to Test Sets for Evaluating Constrained Generative Models,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254125113, title = {Operationalizing Specifications, In Addition to Test Sets for Evaluating Constrained Generative Models}, author = {{Vikas Raunak} and {Matt Post} and {Arul Menezes}}, year = 2022, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/ad2149957cd288a5626adcce48f9981a2ab59184}, } - Zili Huang, Desh Raj, Leibny Paola García-Perera, and S. Khudanpur, “Adapting Self-Supervised Models to Multi-Talker Speech Recognition Using Speaker Embeddings,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{253244355, title = {Adapting Self-Supervised Models to Multi-Talker Speech Recognition Using Speaker Embeddings}, author = {{Zili Huang} and {Desh Raj} and {Leibny Paola García-Perera} and {S. Khudanpur}}, year = 2022, month = {11}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/2226b25c6656e1d7c99667b4e685cd01348e8577}, } - W. G. C. Bandara, Naman Patel, A. Gholami, Mehdi Nikkhah, M. Agrawal, and Vishal M. Patel, “AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoders,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{253553494, title = {AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoders}, author = {{W. G. C. Bandara} and {Naman Patel} and {A. Gholami} and {Mehdi Nikkhah} and {M. Agrawal} and {Vishal M. Patel}}, year = 2022, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/a135632a05cc1f3311859fdebcd1350b4e9e1ee7}, } - Yuanze Lin, Chen Wei, Huiyu Wang, A. Yuille, and Cihang Xie, “SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-training,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{253735003, title = {SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-training}, author = {{Yuanze Lin} and {Chen Wei} and {Huiyu Wang} and {A. Yuille} and {Cihang Xie}}, year = 2022, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/210f6ffbed4bf3a0f020cfcb48dab9d6a9939cdb}, } - Elias Stengel-Eskin and Benjamin Van Durme, “Calibrated Interpretation: Confidence Estimation in Semantic Parsing,” in Transactions of the Association for Computational Linguistics, 2022.
[BibTeX] [Link]@inproceedings{253510101, title = {Calibrated Interpretation: Confidence Estimation in Semantic Parsing}, author = {{Elias Stengel-Eskin} and {Benjamin Van Durme}}, year = 2022, month = {11}, booktitle = {Transactions of the Association for Computational Linguistics}, url = {https://www.semanticscholar.org/paper/c428f1621f79925311082d8d7425dd4d50cd64ed}, } - Dongji Gao, Jiatong Shi, Shun-Po Chuang, Leibny Paola García-Perera, Hung-yi Lee, Shinji Watanabe, and S. Khudanpur, “Euro: Espnet Unsupervised ASR Open-Source Toolkit,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{254095971, title = {Euro: Espnet Unsupervised ASR Open-Source Toolkit}, author = {{Dongji Gao} and {Jiatong Shi} and {Shun-Po Chuang} and {Leibny Paola García-Perera} and {Hung-yi Lee} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2022, month = {11}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/012771aa3a8d59401d22fade9416dbaad22f42b1}, } - Thanh Nguyen-Tang, Ming Yin, S. Gupta, S. Venkatesh, and R. Arora, “On Instance-Dependent Bounds for Offline Reinforcement Learning with Linear Function Approximation,” in AAAI Conference on Artificial Intelligence, 2022.
[BibTeX] [Link]@inproceedings{253801674, title = {On Instance-Dependent Bounds for Offline Reinforcement Learning with Linear Function Approximation}, author = {{Thanh Nguyen-Tang} and {Ming Yin} and {S. Gupta} and {S. Venkatesh} and {R. Arora}}, year = 2022, month = {11}, booktitle = {AAAI Conference on Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/b61a3d718a192e39a437d32a6ed4037b8c29cc41}, } - Yu Zeng, Zhe Lin, Jianming Zhang, Qing Liu, J. Collomosse, Jason Kuen, and Vishal M. Patel, “SceneComposer: Any-Level Semantic Image Synthesis,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{253734941, title = {SceneComposer: Any-Level Semantic Image Synthesis}, author = {{Yu Zeng} and {Zhe Lin} and {Jianming Zhang} and {Qing Liu} and {J. Collomosse} and {Jason Kuen} and {Vishal M. Patel}}, year = 2022, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/4cc5266166478592ec8539a2b940740b8d380cdd}, } - Shuyue Stella Li and Kenton Murray, “Language Agnostic Code-Mixing Data Augmentation by Predicting Linguistic Patterns,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{253510862, title = {Language Agnostic Code-Mixing Data Augmentation by Predicting Linguistic Patterns}, author = {{Shuyue Stella Li} and {Kenton Murray}}, year = 2022, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/96fdfc1ba9588d1fab990d504aa590233216326a}, } - Bardia Safaei, VS Vibashan, Celso M. de Melo, Shuowen Hu, and Vishal M. Patel, “Open-Set Automatic Target Recognition,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{253499210, title = {Open-Set Automatic Target Recognition}, author = {{Bardia Safaei} and {VS Vibashan} and {Celso M. de Melo} and {Shuowen Hu} and {Vishal M. Patel}}, year = 2022, month = {11}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/878d61661e35c80c0b981fe4512fbad6c55ab037}, } - V. Rennoll, I. McLane, Mounya Elhilali, and James E. West, “Optimized Acoustic Phantom Design for Characterizing Body Sound Sensors,” in Italian National Conference on Sensors, 2022.
[BibTeX] [Link]@inproceedings{253903852, title = {Optimized Acoustic Phantom Design for Characterizing Body Sound Sensors}, author = {{V. Rennoll} and {I. McLane} and {Mounya Elhilali} and {James E. West}}, year = 2022, month = {11}, booktitle = {Italian National Conference on Sensors}, url = {https://www.semanticscholar.org/paper/0d7b6b5a15b47c1cd1d688f043fd06ff6822d5a1}, } - Shuyang Sun, Jieneng Chen, Ruifei He, A. Yuille, Philip H. S. Torr, and Song Bai, “LUMix: Improving Mixup by Better Modelling Label Uncertainty,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254069733, title = {LUMix: Improving Mixup by Better Modelling Label Uncertainty}, author = {{Shuyang Sun} and {Jieneng Chen} and {Ruifei He} and {A. Yuille} and {Philip H. S. Torr} and {Song Bai}}, year = 2022, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/62ce349a6dbc58f64ae02d7203c2f9a06cf6f6d4}, } - Hexin Liu, Leibny Paola García-Perera, Andy W. H. Khong, E. Chng, S. Styles, and S. Khudanpur, “Efficient Self-Supervised Learning Representations for Spoken Language Identification,” in IEEE Journal on Selected Topics in Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{251829168, title = {Efficient Self-Supervised Learning Representations for Spoken Language Identification}, author = {{Hexin Liu} and {Leibny Paola García-Perera} and {Andy W. H. Khong} and {E. Chng} and {S. Styles} and {S. Khudanpur}}, year = 2022, month = {10}, booktitle = {IEEE Journal on Selected Topics in Signal Processing}, url = {https://www.semanticscholar.org/paper/130693386f2f7b7c1a98c4298c4ed27b9a56f79e}, } - Junfei Xiao, Yutong Bai, A. Yuille, and Zongwei Zhou, “Delving into Masked Autoencoders for Multi-Label Thorax Disease Classification,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{253098023, title = {Delving into Masked Autoencoders for Multi-Label Thorax Disease Classification}, author = {{Junfei Xiao} and {Yutong Bai} and {A. Yuille} and {Zongwei Zhou}}, year = 2022, month = {10}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/249e00445585586214e27d1f4ade032533132d0a}, } - Qixing Hu, Junfei Xiao, Yixiong Chen, Shuwen Sun, Jieneng Chen, A. Yuille, and Zongwei Zhou, “Synthetic Tumors Make AI Segment Tumors Better,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{253117124, title = {Synthetic Tumors Make AI Segment Tumors Better}, author = {{Qixing Hu} and {Junfei Xiao} and {Yixiong Chen} and {Shuwen Sun} and {Jieneng Chen} and {A. Yuille} and {Zongwei Zhou}}, year = 2022, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/0077f46c9cf3de56319aad65e419131e2466b848}, } - Chenglin Yang, Siyuan Qiao, Qihang Yu, Xiaoding Yuan, Yukun Zhu, A. Yuille, Hartwig Adam, and Liang-Chieh Chen, “MOAT: Alternating Mobile Convolution and Attention Brings Strong Vision Models,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{252715598, title = {MOAT: Alternating Mobile Convolution and Attention Brings Strong Vision Models}, author = {{Chenglin Yang} and {Siyuan Qiao} and {Qihang Yu} and {Xiaoding Yuan} and {Yukun Zhu} and {A. Yuille} and {Hartwig Adam} and {Liang-Chieh Chen}}, year = 2022, month = {10}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/a8a2a8229f99c291bf71ec92b801a073854c52e2}, } - Xiang Xiang, Feng Wang, Yuwen Tan, and A. Yuille, “Imbalanced regression for intensity series of pain expression from videos by regularizing spatio-temporal face nets,” in Pattern Recognition Letters, 2022.
[BibTeX] [Link]@inproceedings{252726978, title = {Imbalanced regression for intensity series of pain expression from videos by regularizing spatio-temporal face nets}, author = {{Xiang Xiang} and {Feng Wang} and {Yuwen Tan} and {A. Yuille}}, year = 2022, month = {10}, booktitle = {Pattern Recognition Letters}, url = {https://www.semanticscholar.org/paper/e9eab79d381d7799e74afd9917e91d47953aa69d}, } - Hexin Liu, Haihua Xu, Leibny Paola Garcia, Andy W. H. Khong, Yi He, and S. Khudanpur, “Reducing Language Confusion for Code-Switching Speech Recognition with Token-Level Language Diarization,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{253116576, title = {Reducing Language Confusion for Code-Switching Speech Recognition with Token-Level Language Diarization}, author = {{Hexin Liu} and {Haihua Xu} and {Leibny Paola Garcia} and {Andy W. H. Khong} and {Yi He} and {S. Khudanpur}}, year = 2022, month = {10}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/1fab5a425ad712bb8245741c5abec00aa80fc123}, } - G. Botev, A. D. McCarthy, W. Wu, and D. Yarowsky, “Deciphering and Characterizing Out-of-Vocabulary Words for Morphologically Rich Languages,” in Proceedings of the 29th International Conference on Computational Linguistics, Gyeongju, Republic of Korea, 2022, p. 5309–5326.
[BibTeX] [Abstract] [Link]
This paper presents a detailed foundational empirical case study of the nature of out-of-vocabulary words encountered in modern text in a moderate-resource language such as Bulgarian, and a multi-faceted distributional analysis of the underlying word-formation processes that can aid in their compositional translation, tagging, parsing, language modeling, and other NLP tasks. Given that out-of-vocabulary (OOV) words generally present a key open challenge to NLP and machine translation systems, especially toward the lower limit of resource availability, there are useful practical insights, as well as corpus-linguistic insights, from both a detailed manual and automatic taxonomic analysis of the types, multidimensional properties, and processing potential for multiple representative OOV data samples.
@inproceedings{botev-etal-2022-deciphering, title = "Deciphering and Characterizing Out-of-Vocabulary Words for Morphologically Rich Languages", author = "Botev, Georgie and McCarthy, Arya D. and Wu, Winston and Yarowsky, David", editor = "Calzolari, Nicoletta and Huang, Chu-Ren and Kim, Hansaem and Pustejovsky, James and Wanner, Leo and Choi, Key-Sun and Ryu, Pum-Mo and Chen, Hsin-Hsi and Donatelli, Lucia and Ji, Heng and Kurohashi, Sadao and Paggio, Patrizia and Xue, Nianwen and Kim, Seokhwan and Hahm, Younggyun and He, Zhong and Lee, Tony Kyungil and Santus, Enrico and Bond, Francis and Na, Seung-Hoon", booktitle = "Proceedings of the 29th International Conference on Computational Linguistics", month = oct, year = "2022", address = "Gyeongju, Republic of Korea", publisher = "International Committee on Computational Linguistics", url = "https://aclanthology.org/2022.coling-1.472", pages = "5309--5326", abstract = "This paper presents a detailed foundational empirical case study of the nature of out-of-vocabulary words encountered in modern text in a moderate-resource language such as Bulgarian, and a multi-faceted distributional analysis of the underlying word-formation processes that can aid in their compositional translation, tagging, parsing, language modeling, and other NLP tasks. Given that out-of-vocabulary (OOV) words generally present a key open challenge to NLP and machine translation systems, especially toward the lower limit of resource availability, there are useful practical insights, as well as corpus-linguistic insights, from both a detailed manual and automatic taxonomic analysis of the types, multidimensional properties, and processing potential for multiple representative OOV data samples.", } - Chris Nalty, Neehar Peri, Joshua Gleason, C. Castillo, Shuowen Hu, T. Bourlai, and R. Chellappa, “A Brief Survey on Person Recognition at a Distance,” in Asilomar Conference on Signals, Systems and Computers, 2022.
[BibTeX] [Link]@inproceedings{254853697, title = {A Brief Survey on Person Recognition at a Distance}, author = {{Chris Nalty} and {Neehar Peri} and {Joshua Gleason} and {C. Castillo} and {Shuowen Hu} and {T. Bourlai} and {R. Chellappa}}, year = 2022, month = {10}, booktitle = {Asilomar Conference on Signals, Systems and Computers}, url = {https://www.semanticscholar.org/paper/6934bd40d21e3bddce5328d29a7e1083e21d0aad}, } - VS Vibashan, Poojan Oza, Vishwanath A. Sindagi, and Vishal M. Patel, “Mixture of Teacher Experts for Source-Free Domain Adaptive Object Detection,” in International Conference on Information Photonics, 2022.
[BibTeX] [Link]@inproceedings{253347117, title = {Mixture of Teacher Experts for Source-Free Domain Adaptive Object Detection}, author = {{VS Vibashan} and {Poojan Oza} and {Vishwanath A. Sindagi} and {Vishal M. Patel}}, year = 2022, month = {10}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/96a609d83a2aaf739fedc4cbfa3f96b14ae234cb}, } - Weiyu Guo, Zhaoshuo Li, Yongkui Yang, Z. Wang, Russell H. Taylor, M. Unberath, A. Yuille, and Yingwei Li, “Context-Enhanced Stereo Transformer,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{253080413, title = {Context-Enhanced Stereo Transformer}, author = {{Weiyu Guo} and {Zhaoshuo Li} and {Yongkui Yang} and {Z. Wang} and {Russell H. Taylor} and {M. Unberath} and {A. Yuille} and {Yingwei Li}}, year = 2022, month = {10}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/3fe123f4777bcb86d796de230b3767c15f28ed7d}, } - J. Zhang, A. DeLucia, and M. Dredze, “Changes in Tweet Geolocation over Time: A Study with Carmen 2.0,” in Proceedings of the Eighth Workshop on Noisy User-generated Text (W-NUT 2022), Gyeongju, Republic of Korea, 2022, p. 1–14.
[BibTeX] [Abstract] [Link]
Researchers across disciplines use Twitter geolocation tools to filter data for desired locations. These tools have largely been trained and tested on English tweets, often originating in the United States from almost a decade ago. Despite the importance of these tools for data curation, the impact of tweet language, country of origin, and creation date on tool performance remains largely unknown. We explore these issues with Carmen, a popular tool for Twitter geolocation. To support this study we introduce Carmen 2.0, a major update which includes the incorporation of GeoNames, a gazetteer that provides much broader coverage of locations. We evaluate using two new Twitter datasets, one for multilingual, multiyear geolocation evaluation, and another for usage trends over time. We found that language, country origin, and time does impact geolocation tool performance.
@inproceedings{zhang-etal-2022-changes, title = "Changes in Tweet Geolocation over Time: A Study with Carmen 2.0", author = "Zhang, Jingyu and DeLucia, Alexandra and Dredze, Mark", booktitle = "Proceedings of the Eighth Workshop on Noisy User-generated Text (W-NUT 2022)", month = oct, year = "2022", address = "Gyeongju, Republic of Korea", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wnut-1.1", pages = "1--14", abstract = "Researchers across disciplines use Twitter geolocation tools to filter data for desired locations. These tools have largely been trained and tested on English tweets, often originating in the United States from almost a decade ago. Despite the importance of these tools for data curation, the impact of tweet language, country of origin, and creation date on tool performance remains largely unknown. We explore these issues with Carmen, a popular tool for Twitter geolocation. To support this study we introduce Carmen 2.0, a major update which includes the incorporation of GeoNames, a gazetteer that provides much broader coverage of locations. We evaluate using two new Twitter datasets, one for multilingual, multiyear geolocation evaluation, and another for usage trends over time. We found that language, country origin, and time does impact geolocation tool performance.", } - Liangyu Chen, Yutong Bai, Siyu Huang, Yongyi Lu, B. Wen, A. Yuille, and Zongwei Zhou, “Making Your First Choice: To Address Cold Start Problem in Vision Active Learning,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{252715847, title = {Making Your First Choice: To Address Cold Start Problem in Vision Active Learning}, author = {{Liangyu Chen} and {Yutong Bai} and {Siyu Huang} and {Yongyi Lu} and {B. Wen} and {A. Yuille} and {Zongwei Zhou}}, year = 2022, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/a4af00f50f0b397b14ae5dc22e0e766c31adaaa8}, } - Kate Sanders, Reno Kriz, Anqi Liu, and Benjamin Van Durme, “Ambiguous Images With Human Judgments for Robust Visual Event Classification,” in Neural Information Processing Systems, 2022.
[BibTeX] [Link]@inproceedings{252735237, title = {Ambiguous Images With Human Judgments for Robust Visual Event Classification}, author = {{Kate Sanders} and {Reno Kriz} and {Anqi Liu} and {Benjamin Van Durme}}, year = 2022, month = {10}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/2a55f57716576fdd5840252d673aabe9a676fced}, } - S. Welleck, Ximing Lu, Peter West, Faeze Brahman, T. Shen, Daniel Khashabi, and Yejin Choi, “Generating Sequences by Learning to Self-Correct,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{253244506, title = {Generating Sequences by Learning to Self-Correct}, author = {{S. Welleck} and {Ximing Lu} and {Peter West} and {Faeze Brahman} and {T. Shen} and {Daniel Khashabi} and {Yejin Choi}}, year = 2022, month = {10}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/538288d24bdad73d831dfed44b706958287ed318}, } - P. Xia and B. Van Durme, “Online Neural Coreference Resolution with Rollback,” in Proceedings of the Fifth Workshop on Computational Models of Reference, Anaphora and Coreference, Gyeongju, Republic of Korea, 2022, p. 13–21.
[BibTeX] [Abstract] [Link]
Humans process natural language online, whether reading a document or participating in multiparty dialogue. Recent advances in neural coreference resolution have focused on offline approaches that assume the full communication history as input. This is neither realistic nor sufficient if we wish to support dialogue understanding in real-time. We benchmark two existing, offline, models and highlight their shortcomings in the online setting. We then modify these models to perform online inference and introduce rollback: a short-term mechanism to correct mistakes. We demonstrate across five English datasets the effectiveness of this approach against an offline and a naive online model in terms of latency, final document-level coreference F1, and average running F1.
@inproceedings{xia-van-durme-2022-online, title = "Online Neural Coreference Resolution with Rollback", author = "Xia, Patrick and Van Durme, Benjamin", editor = "Ogrodniczuk, Maciej and Pradhan, Sameer and Nedoluzhko, Anna and Ng, Vincent and Poesio, Massimo", booktitle = "Proceedings of the Fifth Workshop on Computational Models of Reference, Anaphora and Coreference", month = oct, year = "2022", address = "Gyeongju, Republic of Korea", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.crac-1.2", pages = "13--21", abstract = "Humans process natural language online, whether reading a document or participating in multiparty dialogue. Recent advances in neural coreference resolution have focused on offline approaches that assume the full communication history as input. This is neither realistic nor sufficient if we wish to support dialogue understanding in real-time. We benchmark two existing, offline, models and highlight their shortcomings in the online setting. We then modify these models to perform online inference and introduce rollback: a short-term mechanism to correct mistakes. We demonstrate across five English datasets the effectiveness of this approach against an offline and a naive online model in terms of latency, final document-level coreference F1, and average running F1.", } - Junfei Xiao, Zhichao Xu, Shiyi Lan, Zhiding Yu, A. Yuille, and Anima Anandkumar, “1st Place Solution of The Robust Vision Challenge (RVC) 2022 Semantic Segmentation Track,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{253098673, title = {1st Place Solution of The Robust Vision Challenge (RVC) 2022 Semantic Segmentation Track}, author = {{Junfei Xiao} and {Zhichao Xu} and {Shiyi Lan} and {Zhiding Yu} and {A. Yuille} and {Anima Anandkumar}}, year = 2022, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/17a6bee0ef616822d8a883f6bc373dd676242793}, } - Shota Horiguchi, Yuki Takashima, Shinji Watanabe, and Leibny Paola García-Perera, “Mutual Learning of Single- and Multi-Channel End-to-End Neural Diarization,” in Spoken Language Technology Workshop, 2022.
[BibTeX] [Link]@inproceedings{252762304, title = {Mutual Learning of Single- and Multi-Channel End-to-End Neural Diarization}, author = {{Shota Horiguchi} and {Yuki Takashima} and {Shinji Watanabe} and {Leibny Paola García-Perera}}, year = 2022, month = {10}, booktitle = {Spoken Language Technology Workshop}, url = {https://www.semanticscholar.org/paper/30472f3386177fb929a8454cbbb70462e30d9c61}, } - Yuxiang Guo, Cheng Peng, Chun Pong Lau, and R. Chellappa, “Multi-Modal Human Authentication Using Silhouettes, Gait and RGB,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2022.
[BibTeX] [Link]@inproceedings{252780362, title = {Multi-Modal Human Authentication Using Silhouettes, Gait and RGB}, author = {{Yuxiang Guo} and {Cheng Peng} and {Chun Pong Lau} and {R. Chellappa}}, year = 2022, month = {10}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/e89d9b5c7b5d9c4b490ba1d5fdbbca423920c3e1}, } - D. Licht, C. Gao, J. Lam, F. Guzman, M. Diab, and P. Koehn, “Consistent Human Evaluation of Machine Translation across Language Pairs,” in Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track), Orlando, USA, 2022, p. 309–321.
[BibTeX] [Abstract] [Link]
Obtaining meaningful quality scores for machine translation systems through human evaluation remains a challenge given the high variability between human evaluators, partly due to subjective expectations for translation quality for different language pairs. We propose a new metric called XSTS that is more focused on semantic equivalence and a cross-lingual calibration method that enables more consistent assessment. We demonstrate the effectiveness of these novel contributions in large scale evaluation studies across up to 14 language pairs, with translation both into and out of English.
@inproceedings{licht-etal-2022-consistent, title = "Consistent Human Evaluation of Machine Translation across Language Pairs", author = "Licht, Daniel and Gao, Cynthia and Lam, Janice and Guzman, Francisco and Diab, Mona and Koehn, Philipp", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track)", month = sep, year = "2022", address = "Orlando, USA", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2022.amta-research.24", pages = "309--321", abstract = "Obtaining meaningful quality scores for machine translation systems through human evaluation remains a challenge given the high variability between human evaluators, partly due to subjective expectations for translation quality for different language pairs. We propose a new metric called XSTS that is more focused on semantic equivalence and a cross-lingual calibration method that enables more consistent assessment. We demonstrate the effectiveness of these novel contributions in large scale evaluation studies across up to 14 language pairs, with translation both into and out of English.", } - N. Verma, K. Murray, and K. Duh, “Strategies for Adapting Multilingual Pre-training for Domain-Specific Machine Translation,” in Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track), Orlando, USA, 2022, p. 31–44.
[BibTeX] [Abstract] [Link]
Pretrained multilingual sequence-to-sequence models have been successful in improving translation performance for mid- and lower-resourced languages. However, it is unclear if these models are helpful in the domain adaptation setting, and if so, how to best adapt them to both the domain and translation language pair. Therefore, in this work, we propose two major fine-tuning strategies: our language-first approach first learns the translation language pair via general bitext, followed by the domain via in-domain bitext, and our domain-first approach first learns the domain via multilingual in-domain bitext, followed by the language pair via language pair-specific in-domain bitext. We test our approach on 3 domains at different levels of data availability, and 5 language pairs. We find that models using an mBART initialization generally outperform those using a random Transformer initialization. This holds for languages even outside of mBART{‘}s pretraining set, and can result in improvements of over +10 BLEU. Additionally, we find that via our domain-first approach, fine-tuning across multilingual in-domain corpora can lead to stark improvements in domain adaptation without sourcing additional out-of-domain bitext. In larger domain availability settings, our domain-first approach can be competitive with our language-first approach, even when using over 50X less data.
@inproceedings{verma-etal-2022-strategies, title = "Strategies for Adapting Multilingual Pre-training for Domain-Specific Machine Translation", author = "Verma, Neha and Murray, Kenton and Duh, Kevin", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track)", month = sep, year = "2022", address = "Orlando, USA", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2022.amta-research.3", pages = "31--44", abstract = "Pretrained multilingual sequence-to-sequence models have been successful in improving translation performance for mid- and lower-resourced languages. However, it is unclear if these models are helpful in the domain adaptation setting, and if so, how to best adapt them to both the domain and translation language pair. Therefore, in this work, we propose two major fine-tuning strategies: our language-first approach first learns the translation language pair via general bitext, followed by the domain via in-domain bitext, and our domain-first approach first learns the domain via multilingual in-domain bitext, followed by the language pair via language pair-specific in-domain bitext. We test our approach on 3 domains at different levels of data availability, and 5 language pairs. We find that models using an mBART initialization generally outperform those using a random Transformer initialization. This holds for languages even outside of mBART{'}s pretraining set, and can result in improvements of over +10 BLEU. Additionally, we find that via our domain-first approach, fine-tuning across multilingual in-domain corpora can lead to stark improvements in domain adaptation without sourcing additional out-of-domain bitext. In larger domain availability settings, our domain-first approach can be competitive with our language-first approach, even when using over 50X less data.", } - S. Sia and K. Duh, “Prefix Embeddings for In-context Machine Translation,” in Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track), Orlando, USA, 2022, p. 45–57.
[BibTeX] [Abstract] [Link]
Very large language models have been shown to translate with few-shot in-context examples. However, they have not achieved state-of-art results for translating out of English. In this work, we investigate an extremely lightweight fixed-parameter method for conditioning a large language model to better translate into the target language. Our method introduces additional embeddings, known as prefix embeddings which do not interfere with the existing weights of the model. Using unsupervised and weakly semi-supervised methods that train only 0.0001{\%} of the model parameters, the simple method improves {\textasciitilde}0.2-1.3 BLEU points across 3 domains and 3 languages. We analyze the resulting embeddings{‘} training dynamics, and where they lie in the embedding space, and show that our trained embeddings can be used for both in-context translation, and diverse generation of the target sentence.
@inproceedings{sia-duh-2022-prefix, title = "Prefix Embeddings for In-context Machine Translation", author = "Sia, Suzanna and Duh, Kevin", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track)", month = sep, year = "2022", address = "Orlando, USA", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2022.amta-research.4", pages = "45--57", abstract = "Very large language models have been shown to translate with few-shot in-context examples. However, they have not achieved state-of-art results for translating out of English. In this work, we investigate an extremely lightweight fixed-parameter method for conditioning a large language model to better translate into the target language. Our method introduces additional embeddings, known as prefix embeddings which do not interfere with the existing weights of the model. Using unsupervised and weakly semi-supervised methods that train only 0.0001{\%} of the model parameters, the simple method improves {\textasciitilde}0.2-1.3 BLEU points across 3 domains and 3 languages. We analyze the resulting embeddings{'} training dynamics, and where they lie in the embedding space, and show that our trained embeddings can be used for both in-context translation, and diverse generation of the target sentence.", } - W. Tan, S. Ding, H. Khayrallah, and P. Koehn, “Doubly-Trained Adversarial Data Augmentation for Neural Machine Translation,” in Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track), Orlando, USA, 2022, p. 157–174.
[BibTeX] [Abstract] [Link]
Neural Machine Translation (NMT) models are known to suffer from noisy inputs. To make models robust, we generate adversarial augmentation samples that attack the model and preserve the source-side meaning at the same time. To generate such samples, we propose a doubly-trained architecture that pairs two NMT models of opposite translation directions with a joint loss function, which combines the target-side attack and the source-side semantic similarity constraint. The results from our experiments across three different language pairs and two evaluation metrics show that these adversarial samples improve model robustness.
@inproceedings{tan-etal-2022-doubly, title = "Doubly-Trained Adversarial Data Augmentation for Neural Machine Translation", author = "Tan, Weiting and Ding, Shuoyang and Khayrallah, Huda and Koehn, Philipp", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of the 15th biennial conference of the Association for Machine Translation in the Americas (Volume 1: Research Track)", month = sep, year = "2022", address = "Orlando, USA", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2022.amta-research.12", pages = "157--174", abstract = "Neural Machine Translation (NMT) models are known to suffer from noisy inputs. To make models robust, we generate adversarial augmentation samples that attack the model and preserve the source-side meaning at the same time. To generate such samples, we propose a doubly-trained architecture that pairs two NMT models of opposite translation directions with a joint loss function, which combines the target-side attack and the source-side semantic similarity constraint. The results from our experiments across three different language pairs and two evaluation metrics show that these adversarial samples improve model robustness.", } - A. Blair-stanek and B. Van Durme, “Improved Induction of Narrative Chains via Cross-Document Relations,” in Proceedings of the 11th Joint Conference on Lexical and Computational Semantics, Seattle, Washington, 2022, p. 208–212. doi:10.18653/v1/2022.starsem-1.18
[BibTeX] [Abstract] [Link]
The standard approach for inducing narrative chains considers statistics gathered per individual document. We consider whether statistics gathered using cross-document relations can lead to improved chain induction. Our study is motivated by legal narratives, where cases typically cite thematically similar cases. We consider four novel variations on pointwise mutual information (PMI), each accounting for cross-document relations in a different way. One proposed PMI variation performs 58{\%} better relative to standard PMI on recall@50 and induces qualitatively better narrative chains.
@inproceedings{blair-stanek-van-durme-2022-improved, title = "Improved Induction of Narrative Chains via Cross-Document Relations", author = "Blair-stanek, Andrew and Van Durme, Benjamin", editor = "Nastase, Vivi and Pavlick, Ellie and Pilehvar, Mohammad Taher and Camacho-Collados, Jose and Raganato, Alessandro", booktitle = "Proceedings of the 11th Joint Conference on Lexical and Computational Semantics", month = jul, year = "2022", address = "Seattle, Washington", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.starsem-1.18", doi = "10.18653/v1/2022.starsem-1.18", pages = "208--212", abstract = "The standard approach for inducing narrative chains considers statistics gathered per individual document. We consider whether statistics gathered using cross-document relations can lead to improved chain induction. Our study is motivated by legal narratives, where cases typically cite thematically similar cases. We consider four novel variations on pointwise mutual information (PMI), each accounting for cross-document relations in a different way. One proposed PMI variation performs 58{\%} better relative to standard PMI on recall@50 and induces qualitatively better narrative chains.", } - R. Volum, S. Rao, M. Xu, G. DesGarennes, C. Brockett, B. Van Durme, O. Deng, A. Malhotra, and B. Dolan, “Craft an Iron Sword: Dynamically Generating Interactive Game Characters by Prompting Large Language Models Tuned on Code,” in Proceedings of the 3rd Wordplay: When Language Meets Games Workshop (Wordplay 2022), Seattle, United States, 2022, p. 25–43. doi:10.18653/v1/2022.wordplay-1.3
[BibTeX] [Abstract] [Link]
Non-Player Characters (NPCs) significantly enhance the player experience in many games. Historically, players{‘} interactions with NPCs have tended to be highly scripted, to be limited to natural language responses to be selected by the player, and to not involve dynamic change in game state. In this work, we demonstrate that use of a few example conversational prompts can power a conversational agent to generate both natural language and novel code. This approach can permit development of NPCs with which players can have grounded conversations that are free-form and less repetitive. We demonstrate our approach using OpenAI Codex (GPT-3 finetuned on GitHub), with Minecraft game development as our test bed. We show that with a few example prompts, a Codex-based agent can generate novel code, hold multi-turn conversations and answer questions about structured data. We evaluate this application using experienced gamers in a Minecraft realm and provide analysis of failure cases and suggest possible directions for solutions.
@inproceedings{volum-etal-2022-craft, title = "Craft an Iron Sword: Dynamically Generating Interactive Game Characters by Prompting Large Language Models Tuned on Code", author = "Volum, Ryan and Rao, Sudha and Xu, Michael and DesGarennes, Gabriel and Brockett, Chris and Van Durme, Benjamin and Deng, Olivia and Malhotra, Akanksha and Dolan, Bill", editor = "C{\^o}t{\'e}, Marc-Alexandre and Yuan, Xingdi and Ammanabrolu, Prithviraj", booktitle = "Proceedings of the 3rd Wordplay: When Language Meets Games Workshop (Wordplay 2022)", month = jul, year = "2022", address = "Seattle, United States", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.wordplay-1.3", doi = "10.18653/v1/2022.wordplay-1.3", pages = "25--43", abstract = "Non-Player Characters (NPCs) significantly enhance the player experience in many games. Historically, players{'} interactions with NPCs have tended to be highly scripted, to be limited to natural language responses to be selected by the player, and to not involve dynamic change in game state. In this work, we demonstrate that use of a few example conversational prompts can power a conversational agent to generate both natural language and novel code. This approach can permit development of NPCs with which players can have grounded conversations that are free-form and less repetitive. We demonstrate our approach using OpenAI Codex (GPT-3 finetuned on GitHub), with Minecraft game development as our test bed. We show that with a few example prompts, a Codex-based agent can generate novel code, hold multi-turn conversations and answer questions about structured data. We evaluate this application using experienced gamers in a Minecraft realm and provide analysis of failure cases and suggest possible directions for solutions.", } - A. Zirikly and M. Dredze, “Explaining Models of Mental Health via Clinically Grounded Auxiliary Tasks,” in Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology, Seattle, USA, 2022, p. 30–39. doi:10.18653/v1/2022.clpsych-1.3
[BibTeX] [Abstract] [Link]
Models of mental health based on natural language processing can uncover latent signals of mental health from language. Models that indicate whether an individual is depressed, or has other mental health conditions, can aid in diagnosis and treatment. A critical aspect of integration of these models into the clinical setting relies on explaining their behavior to domain experts. In the case of mental health diagnosis, clinicians already rely on an assessment framework to make these decisions; that framework can help a model generate meaningful explanations. In this work we propose to use PHQ-9 categories as an auxiliary task to explaining a social media based model of depression. We develop a multi-task learning framework that predicts both depression and PHQ-9 categories as auxiliary tasks. We compare the quality of explanations generated based on the depression task only, versus those that use the predicted PHQ-9 categories. We find that by relying on clinically meaningful auxiliary tasks, we produce more meaningful explanations.
@inproceedings{zirikly-dredze-2022-explaining, title = "Explaining Models of Mental Health via Clinically Grounded Auxiliary Tasks", author = "Zirikly, Ayah and Dredze, Mark", editor = "Zirikly, Ayah and Atzil-Slonim, Dana and Liakata, Maria and Bedrick, Steven and Desmet, Bart and Ireland, Molly and Lee, Andrew and MacAvaney, Sean and Purver, Matthew and Resnik, Rebecca and Yates, Andrew", booktitle = "Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology", month = jul, year = "2022", address = "Seattle, USA", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.clpsych-1.3", doi = "10.18653/v1/2022.clpsych-1.3", pages = "30--39", abstract = "Models of mental health based on natural language processing can uncover latent signals of mental health from language. Models that indicate whether an individual is depressed, or has other mental health conditions, can aid in diagnosis and treatment. A critical aspect of integration of these models into the clinical setting relies on explaining their behavior to domain experts. In the case of mental health diagnosis, clinicians already rely on an assessment framework to make these decisions; that framework can help a model generate meaningful explanations. In this work we propose to use PHQ-9 categories as an auxiliary task to explaining a social media based model of depression. We develop a multi-task learning framework that predicts both depression and PHQ-9 categories as auxiliary tasks. We compare the quality of explanations generated based on the depression task only, versus those that use the predicted PHQ-9 categories. We find that by relying on clinically meaningful auxiliary tasks, we produce more meaningful explanations.", } - A. Tsakalidis, J. Chim, I. M. Bilal, A. Zirikly, D. Atzil-Slonim, F. Nanni, P. Resnik, M. Gaur, K. Roy, B. Inkster, J. Leintz, and M. Liakata, “Overview of the CLPsych 2022 Shared Task: Capturing Moments of Change in Longitudinal User Posts,” in Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology, Seattle, USA, 2022, p. 184–198. doi:10.18653/v1/2022.clpsych-1.16
[BibTeX] [Abstract] [Link]
We provide an overview of the CLPsych 2022 Shared Task, which focusses on the automatic identification of {`}Moments of Change{‘} in lon- gitudinal posts by individuals on social media and its connection with information regarding mental health . This year{‘}s task introduced the notion of longitudinal modelling of the text generated by an individual online over time, along with appropriate temporally sen- sitive evaluation metrics. The Shared Task con- sisted of two subtasks: (a) the main task of cap- turing changes in an individual{‘}s mood (dras- tic changes-{`}Switches{‘}- and gradual changes -{`}Escalations{‘}- on the basis of textual content shared online; and subsequently (b) the sub- task of identifying the suicide risk level of an individual {–} a continuation of the CLPsych 2019 Shared Task{–} where participants were encouraged to explore how the identification of changes in mood in task (a) can help with assessing suicidality risk in task (b).
@inproceedings{tsakalidis-etal-2022-overview, title = "Overview of the {CLP}sych 2022 Shared Task: Capturing Moments of Change in Longitudinal User Posts", author = "Tsakalidis, Adam and Chim, Jenny and Bilal, Iman Munire and Zirikly, Ayah and Atzil-Slonim, Dana and Nanni, Federico and Resnik, Philip and Gaur, Manas and Roy, Kaushik and Inkster, Becky and Leintz, Jeff and Liakata, Maria", editor = "Zirikly, Ayah and Atzil-Slonim, Dana and Liakata, Maria and Bedrick, Steven and Desmet, Bart and Ireland, Molly and Lee, Andrew and MacAvaney, Sean and Purver, Matthew and Resnik, Rebecca and Yates, Andrew", booktitle = "Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology", month = jul, year = "2022", address = "Seattle, USA", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.clpsych-1.16", doi = "10.18653/v1/2022.clpsych-1.16", pages = "184--198", abstract = "We provide an overview of the CLPsych 2022 Shared Task, which focusses on the automatic identification of {`}Moments of Change{'} in lon- gitudinal posts by individuals on social media and its connection with information regarding mental health . This year{'}s task introduced the notion of longitudinal modelling of the text generated by an individual online over time, along with appropriate temporally sen- sitive evaluation metrics. The Shared Task con- sisted of two subtasks: (a) the main task of cap- turing changes in an individual{'}s mood (dras- tic changes-{`}Switches{'}- and gradual changes -{`}Escalations{'}- on the basis of textual content shared online; and subsequently (b) the sub- task of identifying the suicide risk level of an individual {--} a continuation of the CLPsych 2019 Shared Task{--} where participants were encouraged to explore how the identification of changes in mood in task (a) can help with assessing suicidality risk in task (b).", } - C. Zhang, B. Van Durme, Z. Li, and E. Stengel-Eskin, “Visual Commonsense in Pretrained Unimodal and Multimodal Models,” in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Seattle, United States, 2022, p. 5321–5335. doi:10.18653/v1/2022.naacl-main.390
[BibTeX] [Abstract] [Link]
Our commonsense knowledge about objects includes their typical visual attributes; we know that bananas are typically yellow or green, and not purple. Text and image corpora, being subject to reporting bias, represent this world-knowledge to varying degrees of faithfulness. In this paper, we investigate to what degree unimodal (language-only) and multimodal (image and language) models capture a broad range of visually salient attributes. To that end, we create the Visual Commonsense Tests (ViComTe) dataset covering 5 property types (color, shape, material, size, and visual co-occurrence) for over 5000 subjects. We validate this dataset by showing that our grounded color data correlates much better than ungrounded text-only data with crowdsourced color judgments provided by Paik et al. (2021). We then use our dataset to evaluate pretrained unimodal models and multimodal models. Our results indicate that multimodal models better reconstruct attribute distributions, but are still subject to reporting bias. Moreover, increasing model size does not enhance performance, suggesting that the key to visual commonsense lies in the data.
@inproceedings{zhang-etal-2022-visual, title = "Visual Commonsense in Pretrained Unimodal and Multimodal Models", author = "Zhang, Chenyu and Van Durme, Benjamin and Li, Zhuowan and Stengel-Eskin, Elias", editor = "Carpuat, Marine and de Marneffe, Marie-Catherine and Meza Ruiz, Ivan Vladimir", booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies", month = jul, year = "2022", address = "Seattle, United States", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.naacl-main.390", doi = "10.18653/v1/2022.naacl-main.390", pages = "5321--5335", abstract = "Our commonsense knowledge about objects includes their typical visual attributes; we know that bananas are typically yellow or green, and not purple. Text and image corpora, being subject to reporting bias, represent this world-knowledge to varying degrees of faithfulness. In this paper, we investigate to what degree unimodal (language-only) and multimodal (image and language) models capture a broad range of visually salient attributes. To that end, we create the Visual Commonsense Tests (ViComTe) dataset covering 5 property types (color, shape, material, size, and visual co-occurrence) for over 5000 subjects. We validate this dataset by showing that our grounded color data correlates much better than ungrounded text-only data with crowdsourced color judgments provided by Paik et al. (2021). We then use our dataset to evaluate pretrained unimodal models and multimodal models. Our results indicate that multimodal models better reconstruct attribute distributions, but are still subject to reporting bias. Moreover, increasing model size does not enhance performance, suggesting that the key to visual commonsense lies in the data.", } - O. Weller, M. Marone, V. Braverman, D. Lawrie, and B. Van Durme, “Pretrained Models for Multilingual Federated Learning,” in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Seattle, United States, 2022, p. 1413–1421. doi:10.18653/v1/2022.naacl-main.101
[BibTeX] [Abstract] [Link]
Since the advent of Federated Learning (FL), research has applied these methods to natural language processing (NLP) tasks. Despite a plethora of papers in FL for NLP, no previous works have studied how multilingual text impacts FL algorithms. Furthermore, multilingual text provides an interesting avenue to examine the impact of non-IID text (e.g. different languages) on FL in naturally occurring data. We explore three multilingual language tasks, language modeling, machine translation, and text classification using differing federated and non-federated learning algorithms. Our results show that using pretrained models reduces the negative effects of FL, helping them to perform near or better than centralized (no privacy) learning, even when using non-IID partitioning.
@inproceedings{weller-etal-2022-pretrained, title = "Pretrained Models for Multilingual Federated Learning", author = "Weller, Orion and Marone, Marc and Braverman, Vladimir and Lawrie, Dawn and Van Durme, Benjamin", editor = "Carpuat, Marine and de Marneffe, Marie-Catherine and Meza Ruiz, Ivan Vladimir", booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies", month = jul, year = "2022", address = "Seattle, United States", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.naacl-main.101", doi = "10.18653/v1/2022.naacl-main.101", pages = "1413--1421", abstract = "Since the advent of Federated Learning (FL), research has applied these methods to natural language processing (NLP) tasks. Despite a plethora of papers in FL for NLP, no previous works have studied how multilingual text impacts FL algorithms. Furthermore, multilingual text provides an interesting avenue to examine the impact of non-IID text (e.g. different languages) on FL in naturally occurring data. We explore three multilingual language tasks, language modeling, machine translation, and text classification using differing federated and non-federated learning algorithms. Our results show that using pretrained models reduces the negative effects of FL, helping them to perform near or better than centralized (no privacy) learning, even when using non-IID partitioning.", } - R. Shin and B. Van Durme, “Few-Shot Semantic Parsing with Language Models Trained on Code,” in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Seattle, United States, 2022, p. 5417–5425. doi:10.18653/v1/2022.naacl-main.396
[BibTeX] [Abstract] [Link]
Large language models can perform semantic parsing with little training data, when prompted with in-context examples. It has been shown that this can be improved by formulating the problem as paraphrasing into canonical utterances, which casts the underlying meaning representation into a controlled natural language-like representation. Intuitively, such models can more easily output canonical utterances as they are closer to the natural language used for pre-training. Recently, models also pre-trained on code, like OpenAI Codex, have risen in prominence. For semantic parsing tasks where we map natural language into code, such models may prove more adept at it. In this paper, we test this hypothesis and find that Codex performs better on such tasks than equivalent GPT-3 models. We evaluate on Overnight and SMCalFlow and find that unlike GPT-3, Codex performs similarly when targeting meaning representations directly, perhaps because meaning representations are structured similar to code in these datasets.
@inproceedings{shin-van-durme-2022-shot, title = "Few-Shot Semantic Parsing with Language Models Trained on Code", author = "Shin, Richard and Van Durme, Benjamin", editor = "Carpuat, Marine and de Marneffe, Marie-Catherine and Meza Ruiz, Ivan Vladimir", booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies", month = jul, year = "2022", address = "Seattle, United States", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.naacl-main.396", doi = "10.18653/v1/2022.naacl-main.396", pages = "5417--5425", abstract = "Large language models can perform semantic parsing with little training data, when prompted with in-context examples. It has been shown that this can be improved by formulating the problem as paraphrasing into canonical utterances, which casts the underlying meaning representation into a controlled natural language-like representation. Intuitively, such models can more easily output canonical utterances as they are closer to the natural language used for pre-training. Recently, models also pre-trained on code, like OpenAI Codex, have risen in prominence. For semantic parsing tasks where we map natural language into code, such models may prove more adept at it. In this paper, we test this hypothesis and find that Codex performs better on such tasks than equivalent GPT-3 models. We evaluate on Overnight and SMCalFlow and find that unlike GPT-3, Codex performs similarly when targeting meaning representations directly, perhaps because meaning representations are structured similar to code in these datasets.", } - A. Zirikly, B. Desmet, J. Porcino, J. Camacho Maldonado, P. Ho, R. Jimenez Silva, and M. Sacco, “A Whole-Person Function Dictionary for the Mobility, Self-Care and Domestic Life Domains: a Seedset Expansion Approach,” in Proceedings of the Thirteenth Language Resources and Evaluation Conference, Marseille, France, 2022, p. 2850–2855.
[BibTeX] [Abstract] [Link]
Whole-person functional limitations in the areas of mobility, self-care and domestic life affect a majority of individuals with disabilities. Detecting, recording and monitoring such limitations would benefit those individuals, as well as research on whole-person functioning and general public health. Dictionaries of terms related to whole-person function would enable automated identification and extraction of relevant information. However, no such terminologies currently exist, due in part to a lack of standardized coding and their availability mainly in free text clinical notes. In this paper, we introduce terminologies of whole-person function in the domains of mobility, self-care and domestic life, built and evaluated using a small set of manually annotated clinical notes, which provided a seedset that was expanded using a mix of lexical and deep learning approaches.
@inproceedings{zirikly-etal-2022-whole, title = "A Whole-Person Function Dictionary for the Mobility, Self-Care and Domestic Life Domains: a Seedset Expansion Approach", author = "Zirikly, Ayah and Desmet, Bart and Porcino, Julia and Camacho Maldonado, Jonathan and Ho, Pei-Shu and Jimenez Silva, Rafael and Sacco, Maryanne", editor = "Calzolari, Nicoletta and B{\'e}chet, Fr{\'e}d{\'e}ric and Blache, Philippe and Choukri, Khalid and Cieri, Christopher and Declerck, Thierry and Goggi, Sara and Isahara, Hitoshi and Maegaard, Bente and Mariani, Joseph and Mazo, H{\'e}l{\`e}ne and Odijk, Jan and Piperidis, Stelios", booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.lrec-1.305", pages = "2850--2855", abstract = "Whole-person functional limitations in the areas of mobility, self-care and domestic life affect a majority of individuals with disabilities. Detecting, recording and monitoring such limitations would benefit those individuals, as well as research on whole-person functioning and general public health. Dictionaries of terms related to whole-person function would enable automated identification and extraction of relevant information. However, no such terminologies currently exist, due in part to a lack of standardized coding and their availability mainly in free text clinical notes. In this paper, we introduce terminologies of whole-person function in the domains of mobility, self-care and domestic life, built and evaluated using a small set of manually annotated clinical notes, which provided a seedset that was expanded using a mix of lexical and deep learning approaches.", } - L. Kanashiro Pereira, “Attention-Focused Adversarial Training for Robust Temporal Reasoning,” in Proceedings of the Thirteenth Language Resources and Evaluation Conference, Marseille, France, 2022, p. 7352–7359.
[BibTeX] [Abstract] [Link]
We propose an enhanced adversarial training algorithm for fine-tuning transformer-based language models (i.e., RoBERTa) and apply it to the temporal reasoning task. Current adversarial training approaches for NLP add the adversarial perturbation only to the embedding layer, ignoring the other layers of the model, which might limit the generalization power of adversarial training. Instead, our algorithm searches for the best combination of layers to add the adversarial perturbation. We add the adversarial perturbation to multiple hidden states or attention representations of the model layers. Adding the perturbation to the attention representations performed best in our experiments. Our model can improve performance on several temporal reasoning benchmarks, and establishes new state-of-the-art results.
@inproceedings{kanashiro-pereira-2022-attention, title = "Attention-Focused Adversarial Training for Robust Temporal Reasoning", author = "Kanashiro Pereira, Lis", editor = "Calzolari, Nicoletta and B{\'e}chet, Fr{\'e}d{\'e}ric and Blache, Philippe and Choukri, Khalid and Cieri, Christopher and Declerck, Thierry and Goggi, Sara and Isahara, Hitoshi and Maegaard, Bente and Mariani, Joseph and Mazo, H{\'e}l{\`e}ne and Odijk, Jan and Piperidis, Stelios", booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.lrec-1.800", pages = "7352--7359", abstract = "We propose an enhanced adversarial training algorithm for fine-tuning transformer-based language models (i.e., RoBERTa) and apply it to the temporal reasoning task. Current adversarial training approaches for NLP add the adversarial perturbation only to the embedding layer, ignoring the other layers of the model, which might limit the generalization power of adversarial training. Instead, our algorithm searches for the best combination of layers to add the adversarial perturbation. We add the adversarial perturbation to multiple hidden states or attention representations of the model layers. Adding the perturbation to the attention representations performed best in our experiments. Our model can improve performance on several temporal reasoning benchmarks, and establishes new state-of-the-art results.", } - K. Batsuren, O. Goldman, S. Khalifa, N. Habash, W. Kiera{‘s}, G. Bella, B. Leonard, G. Nicolai, K. Gorman, Y. G. Ate, M. Ryskina, S. Mielke, E. Budianskaya, C. El-Khaissi, T. Pimentel, M. Gasser, W. A. Lane, M. Raj, M. Coler, J. R. M. Samame, D. S. Camaiteri, E. Z. Rojas, D. López Francis, A. Oncevay, J. López Bautista, G. C. S. Villegas, L. T. Hennigen, A. Ek, D. Guriel, P. Dirix, J. Bernardy, A. Scherbakov, A. Bayyr-ool, A. Anastasopoulos, R. Zariquiey, K. Sheifer, S. Ganieva, H. Cruz, R. Karahó{v{g}}a, S. Markantonatou, G. Pavlidis, M. Plugaryov, E. Klyachko, A. Salehi, C. Angulo, J. Baxi, A. Krizhanovsky, N. Krizhanovskaya, E. Salesky, C. Vania, S. Ivanova, J. White, R. H. Maudslay, J. Valvoda, R. Zmigrod, P. Czarnowska, I. Nikkarinen, A. Salchak, B. Bhatt, C. Straughn, Z. Liu, J. N. Washington, Y. Pinter, D. Ataman, M. Wolinski, T. Suhardijanto, A. Yablonskaya, N. Stoehr, H. Dolatian, Z. Nuriah, S. Ratan, F. M. Tyers, E. M. Ponti, G. Aiton, A. Arora, R. J. Hatcher, R. Kumar, J. Young, D. Rodionova, A. Yemelina, T. Andrushko, I. Marchenko, P. Mashkovtseva, A. Serova, E. Prud{‘}hommeaux, M. Nepomniashchaya, F. Giunchiglia, E. Chodroff, M. Hulden, M. Silfverberg, A. D. McCarthy, D. Yarowsky, R. Cotterell, R. Tsarfaty, and E. Vylomova, “UniMorph 4.0: Universal Morphology,” in Proceedings of the Thirteenth Language Resources and Evaluation Conference, Marseille, France, 2022, p. 840–855.
[BibTeX] [Abstract] [Link]
The Universal Morphology (UniMorph) project is a collaborative effort providing broad-coverage instantiated normalized morphological inflection tables for hundreds of diverse world languages. The project comprises two major thrusts: a language-independent feature schema for rich morphological annotation, and a type-level resource of annotated data in diverse languages realizing that schema. This paper presents the expansions and improvements on several fronts that were made in the last couple of years (since McCarthy et al. (2020)). Collaborative efforts by numerous linguists have added 66 new languages, including 24 endangered languages. We have implemented several improvements to the extraction pipeline to tackle some issues, e.g., missing gender and macrons information. We have amended the schema to use a hierarchical structure that is needed for morphological phenomena like multiple-argument agreement and case stacking, while adding some missing morphological features to make the schema more inclusive. In light of the last UniMorph release, we also augmented the database with morpheme segmentation for 16 languages. Lastly, this new release makes a push towards inclusion of derivational morphology in UniMorph by enriching the data and annotation schema with instances representing derivational processes from MorphyNet.
@inproceedings{batsuren-etal-2022-unimorph, title = "{U}ni{M}orph 4.0: {U}niversal {M}orphology", author = "Batsuren, Khuyagbaatar and Goldman, Omer and Khalifa, Salam and Habash, Nizar and Kiera{\'s}, Witold and Bella, G{\'a}bor and Leonard, Brian and Nicolai, Garrett and Gorman, Kyle and Ate, Yustinus Ghanggo and Ryskina, Maria and Mielke, Sabrina and Budianskaya, Elena and El-Khaissi, Charbel and Pimentel, Tiago and Gasser, Michael and Lane, William Abbott and Raj, Mohit and Coler, Matt and Samame, Jaime Rafael Montoya and Camaiteri, Delio Siticonatzi and Rojas, Esa{\'u} Zumaeta and L{\'o}pez Francis, Didier and Oncevay, Arturo and L{\'o}pez Bautista, Juan and Villegas, Gema Celeste Silva and Hennigen, Lucas Torroba and Ek, Adam and Guriel, David and Dirix, Peter and Bernardy, Jean-Philippe and Scherbakov, Andrey and Bayyr-ool, Aziyana and Anastasopoulos, Antonios and Zariquiey, Roberto and Sheifer, Karina and Ganieva, Sofya and Cruz, Hilaria and Karah{\'o}{\v{g}}a, Ritv{\'a}n and Markantonatou, Stella and Pavlidis, George and Plugaryov, Matvey and Klyachko, Elena and Salehi, Ali and Angulo, Candy and Baxi, Jatayu and Krizhanovsky, Andrew and Krizhanovskaya, Natalia and Salesky, Elizabeth and Vania, Clara and Ivanova, Sardana and White, Jennifer and Maudslay, Rowan Hall and Valvoda, Josef and Zmigrod, Ran and Czarnowska, Paula and Nikkarinen, Irene and Salchak, Aelita and Bhatt, Brijesh and Straughn, Christopher and Liu, Zoey and Washington, Jonathan North and Pinter, Yuval and Ataman, Duygu and Wolinski, Marcin and Suhardijanto, Totok and Yablonskaya, Anna and Stoehr, Niklas and Dolatian, Hossep and Nuriah, Zahroh and Ratan, Shyam and Tyers, Francis M. and Ponti, Edoardo M. and Aiton, Grant and Arora, Aryaman and Hatcher, Richard J. and Kumar, Ritesh and Young, Jeremiah and Rodionova, Daria and Yemelina, Anastasia and Andrushko, Taras and Marchenko, Igor and Mashkovtseva, Polina and Serova, Alexandra and Prud{'}hommeaux, Emily and Nepomniashchaya, Maria and Giunchiglia, Fausto and Chodroff, Eleanor and Hulden, Mans and Silfverberg, Miikka and McCarthy, Arya D. and Yarowsky, David and Cotterell, Ryan and Tsarfaty, Reut and Vylomova, Ekaterina", editor = "Calzolari, Nicoletta and B{\'e}chet, Fr{\'e}d{\'e}ric and Blache, Philippe and Choukri, Khalid and Cieri, Christopher and Declerck, Thierry and Goggi, Sara and Isahara, Hitoshi and Maegaard, Bente and Mariani, Joseph and Mazo, H{\'e}l{\`e}ne and Odijk, Jan and Piperidis, Stelios", booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.lrec-1.89", pages = "840--855", abstract = "The Universal Morphology (UniMorph) project is a collaborative effort providing broad-coverage instantiated normalized morphological inflection tables for hundreds of diverse world languages. The project comprises two major thrusts: a language-independent feature schema for rich morphological annotation, and a type-level resource of annotated data in diverse languages realizing that schema. This paper presents the expansions and improvements on several fronts that were made in the last couple of years (since McCarthy et al. (2020)). Collaborative efforts by numerous linguists have added 66 new languages, including 24 endangered languages. We have implemented several improvements to the extraction pipeline to tackle some issues, e.g., missing gender and macrons information. We have amended the schema to use a hierarchical structure that is needed for morphological phenomena like multiple-argument agreement and case stacking, while adding some missing morphological features to make the schema more inclusive. In light of the last UniMorph release, we also augmented the database with morpheme segmentation for 16 languages. Lastly, this new release makes a push towards inclusion of derivational morphology in UniMorph by enriching the data and annotation schema with instances representing derivational processes from MorphyNet.", } - W. Wu and D. Yarowsky, “On the Robustness of Cognate Generation Models,” in Proceedings of the Thirteenth Language Resources and Evaluation Conference, Marseille, France, 2022, p. 4299–4305.
[BibTeX] [Abstract] [Link]
We evaluate two popular neural cognate generation models{‘} robustness to several types of human-plausible noise (deletion, duplication, swapping, and keyboard errors, as well as a new type of error, phonological errors). We find that duplication and phonological substitution is least harmful, while the other types of errors are harmful. We present an in-depth analysis of the models{‘} results with respect to each error type to explain how and why these models perform as they do.
@inproceedings{wu-yarowsky-2022-robustness, title = "On the Robustness of Cognate Generation Models", author = "Wu, Winston and Yarowsky, David", editor = "Calzolari, Nicoletta and B{\'e}chet, Fr{\'e}d{\'e}ric and Blache, Philippe and Choukri, Khalid and Cieri, Christopher and Declerck, Thierry and Goggi, Sara and Isahara, Hitoshi and Maegaard, Bente and Mariani, Joseph and Mazo, H{\'e}l{\`e}ne and Odijk, Jan and Piperidis, Stelios", booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.lrec-1.458", pages = "4299--4305", abstract = "We evaluate two popular neural cognate generation models{'} robustness to several types of human-plausible noise (deletion, duplication, swapping, and keyboard errors, as well as a new type of error, phonological errors). We find that duplication and phonological substitution is least harmful, while the other types of errors are harmful. We present an in-depth analysis of the models{'} results with respect to each error type to explain how and why these models perform as they do.", } - N. Weber, A. Belyy, N. Holzenberger, R. Rudinger, and B. Van Durme, “Human Schema Curation via Causal Association Rule Mining,” in Proceedings of the 16th Linguistic Annotation Workshop (LAW-XVI) within LREC2022, Marseille, France, 2022, p. 139–150.
[BibTeX] [Abstract] [Link]
Event schemas are structured knowledge sources defining typical real-world scenarios (e.g., going to an airport). We present a framework for efficient human-in-the-loop construction of a schema library, based on a novel script induction system and a well-crafted interface that allows non-experts to {“}program{”} complex event structures. Associated with this work we release a schema library: a machine readable resource of 232 detailed event schemas, each of which describe a distinct typical scenario in terms of its relevant sub-event structure (what happens in the scenario), participants (who plays a role in the scenario), fine-grained typing of each participant, and the implied relational constraints between them. We make our schema library and the SchemaBlocks interface available online.
@inproceedings{weber-etal-2022-human, title = "Human Schema Curation via Causal Association Rule Mining", author = "Weber, Noah and Belyy, Anton and Holzenberger, Nils and Rudinger, Rachel and Van Durme, Benjamin", editor = "Pradhan, Sameer and Kuebler, Sandra", booktitle = "Proceedings of the 16th Linguistic Annotation Workshop (LAW-XVI) within LREC2022", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.law-1.17", pages = "139--150", abstract = "Event schemas are structured knowledge sources defining typical real-world scenarios (e.g., going to an airport). We present a framework for efficient human-in-the-loop construction of a schema library, based on a novel script induction system and a well-crafted interface that allows non-experts to {``}program{''} complex event structures. Associated with this work we release a schema library: a machine readable resource of 232 detailed event schemas, each of which describe a distinct typical scenario in terms of its relevant sub-event structure (what happens in the scenario), participants (who plays a role in the scenario), fine-grained typing of each participant, and the implied relational constraints between them. We make our schema library and the SchemaBlocks interface available online.", } - T. Nguyen, A. Yates, A. Zirikly, B. Desmet, and A. Cohan, “Improving the Generalizability of Depression Detection by Leveraging Clinical Questionnaires,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Dublin, Ireland, 2022, p. 8446–8459. doi:10.18653/v1/2022.acl-long.578
[BibTeX] [Abstract] [Link]
Automated methods have been widely used to identify and analyze mental health conditions (e.g., depression) from various sources of information, including social media. Yet, deployment of such models in real-world healthcare applications faces challenges including poor out-of-domain generalization and lack of trust in black box models. In this work, we propose approaches for depression detection that are constrained to different degrees by the presence of symptoms described in PHQ9, a questionnaire used by clinicians in the depression screening process. In dataset-transfer experiments on three social media datasets, we find that grounding the model in PHQ9{‘}s symptoms substantially improves its ability to generalize to out-of-distribution data compared to a standard BERT-based approach. Furthermore, this approach can still perform competitively on in-domain data. These results and our qualitative analyses suggest that grounding model predictions in clinically-relevant symptoms can improve generalizability while producing a model that is easier to inspect.
@inproceedings{nguyen-etal-2022-improving, title = "Improving the Generalizability of Depression Detection by Leveraging Clinical Questionnaires", author = "Nguyen, Thong and Yates, Andrew and Zirikly, Ayah and Desmet, Bart and Cohan, Arman", editor = "Muresan, Smaranda and Nakov, Preslav and Villavicencio, Aline", booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.acl-long.578", doi = "10.18653/v1/2022.acl-long.578", pages = "8446--8459", abstract = "Automated methods have been widely used to identify and analyze mental health conditions (e.g., depression) from various sources of information, including social media. Yet, deployment of such models in real-world healthcare applications faces challenges including poor out-of-domain generalization and lack of trust in black box models. In this work, we propose approaches for depression detection that are constrained to different degrees by the presence of symptoms described in PHQ9, a questionnaire used by clinicians in the depression screening process. In dataset-transfer experiments on three social media datasets, we find that grounding the model in PHQ9{'}s symptoms substantially improves its ability to generalize to out-of-distribution data compared to a standard BERT-based approach. Furthermore, this approach can still perform competitively on in-domain data. These results and our qualitative analyses suggest that grounding model predictions in clinically-relevant symptoms can improve generalizability while producing a model that is easier to inspect.", } - S. Wu, B. Van Durme, and M. Dredze, “Zero-shot Cross-lingual Transfer is Under-specified Optimization,” in Proceedings of the 7th Workshop on Representation Learning for NLP, Dublin, Ireland, 2022, p. 236–248. doi:10.18653/v1/2022.repl4nlp-1.25
[BibTeX] [Abstract] [Link]
Pretrained multilingual encoders enable zero-shot cross-lingual transfer, but often produce unreliable models that exhibit high performance variance on the target language. We postulate that this high variance results from zero-shot cross-lingual transfer solving an under-specified optimization problem. We show that any linear-interpolated model between the source language monolingual model and source + target bilingual model has equally low source language generalization error, yet the target language generalization error reduces smoothly and linearly as we move from the monolingual to bilingual model, suggesting that the model struggles to identify good solutions for both source and target languages using the source language alone. Additionally, we show that zero-shot solution lies in non-flat region of target language error generalization surface, causing the high variance.
@inproceedings{wu-etal-2022-zero, title = "Zero-shot Cross-lingual Transfer is Under-specified Optimization", author = "Wu, Shijie and Van Durme, Benjamin and Dredze, Mark", editor = "Gella, Spandana and He, He and Majumder, Bodhisattwa Prasad and Can, Burcu and Giunchiglia, Eleonora and Cahyawijaya, Samuel and Min, Sewon and Mozes, Maximilian and Li, Xiang Lorraine and Augenstein, Isabelle and Rogers, Anna and Cho, Kyunghyun and Grefenstette, Edward and Rimell, Laura and Dyer, Chris", booktitle = "Proceedings of the 7th Workshop on Representation Learning for NLP", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.repl4nlp-1.25", doi = "10.18653/v1/2022.repl4nlp-1.25", pages = "236--248", abstract = "Pretrained multilingual encoders enable zero-shot cross-lingual transfer, but often produce unreliable models that exhibit high performance variance on the target language. We postulate that this high variance results from zero-shot cross-lingual transfer solving an under-specified optimization problem. We show that any linear-interpolated model between the source language monolingual model and source + target bilingual model has equally low source language generalization error, yet the target language generalization error reduces smoothly and linearly as we move from the monolingual to bilingual model, suggesting that the model struggles to identify good solutions for both source and target languages using the source language alone. Additionally, we show that zero-shot solution lies in non-flat region of target language error generalization surface, causing the high variance.", } - J. Yang, A. Hussein, M. Wiesner, and S. Khudanpur, “JHU IWSLT 2022 Dialect Speech Translation System Description,” in Proceedings of the 19th International Conference on Spoken Language Translation (IWSLT 2022), Dublin, Ireland (in-person and online), 2022, p. 319–326. doi:10.18653/v1/2022.iwslt-1.29
[BibTeX] [Abstract] [Link]
This paper details the Johns Hopkins speech translation (ST) system used in the IWLST2022 dialect speech translation task. Our system uses a cascade of automatic speech recognition (ASR) and machine translation (MT). We use a Conformer model for ASR systems and a Transformer model for machine translation. Surprisingly, we found that while using additional ASR training data resulted in only a negligible change in performance as measured by BLEU or word error rate (WER), aggressive text normalization improved BLEU more significantly. We also describe an approach, similar to back-translation, for improving performance using synthetic dialectal source text produced from source sentences in mismatched dialects.
@inproceedings{yang-etal-2022-jhu, title = "{JHU} {IWSLT} 2022 Dialect Speech Translation System Description", author = "Yang, Jinyi and Hussein, Amir and Wiesner, Matthew and Khudanpur, Sanjeev", editor = "Salesky, Elizabeth and Federico, Marcello and Costa-juss{\`a}, Marta", booktitle = "Proceedings of the 19th International Conference on Spoken Language Translation (IWSLT 2022)", month = may, year = "2022", address = "Dublin, Ireland (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.iwslt-1.29", doi = "10.18653/v1/2022.iwslt-1.29", pages = "319--326", abstract = "This paper details the Johns Hopkins speech translation (ST) system used in the IWLST2022 dialect speech translation task. Our system uses a cascade of automatic speech recognition (ASR) and machine translation (MT). We use a Conformer model for ASR systems and a Transformer model for machine translation. Surprisingly, we found that while using additional ASR training data resulted in only a negligible change in performance as measured by BLEU or word error rate (WER), aggressive text normalization improved BLEU more significantly. We also describe an approach, similar to back-translation, for improving performance using synthetic dialectal source text produced from source sentences in mismatched dialects.", } - A. Belyy, C. Huang, J. Andreas, E. A. Platanios, S. Thomson, R. Shin, S. Roy, A. Nisnevich, C. Chen, and B. Van Durme, “Guided K-best Selection for Semantic Parsing Annotation,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, Dublin, Ireland, 2022, p. 114–126. doi:10.18653/v1/2022.acl-demo.11
[BibTeX] [Abstract] [Link]
Collecting data for conversational semantic parsing is a time-consuming and demanding process. In this paper we consider, given an incomplete dataset with only a small amount of data, how to build an AI-powered human-in-the-loop process to enable efficient data collection. A guided K-best selection process is proposed, which (i) generates a set of possible valid candidates; (ii) allows users to quickly traverse the set and filter incorrect parses; and (iii) asks users to select the correct parse, with minimal modification when necessary. We investigate how to best support users in efficiently traversing the candidate set and locating the correct parse, in terms of speed and accuracy. In our user study, consisting of five annotators labeling 300 instances each, we find that combining keyword searching, where keywords can be used to query relevant candidates, and keyword suggestion, where representative keywords are automatically generated, enables fast and accurate annotation.
@inproceedings{belyy-etal-2022-guided, title = "Guided K-best Selection for Semantic Parsing Annotation", author = "Belyy, Anton and Huang, Chieh-yang and Andreas, Jacob and Platanios, Emmanouil Antonios and Thomson, Sam and Shin, Richard and Roy, Subhro and Nisnevich, Aleksandr and Chen, Charles and Van Durme, Benjamin", editor = "Basile, Valerio and Kozareva, Zornitsa and Stajner, Sanja", booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: System Demonstrations", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.acl-demo.11", doi = "10.18653/v1/2022.acl-demo.11", pages = "114--126", abstract = "Collecting data for conversational semantic parsing is a time-consuming and demanding process. In this paper we consider, given an incomplete dataset with only a small amount of data, how to build an AI-powered human-in-the-loop process to enable efficient data collection. A guided K-best selection process is proposed, which (i) generates a set of possible valid candidates; (ii) allows users to quickly traverse the set and filter incorrect parses; and (iii) asks users to select the correct parse, with minimal modification when necessary. We investigate how to best support users in efficiently traversing the candidate set and locating the correct parse, in terms of speed and accuracy. In our user study, consisting of five annotators labeling 300 instances each, we find that combining keyword searching, where keywords can be used to query relevant candidates, and keyword suggestion, where representative keywords are automatically generated, enables fast and accurate annotation.", } - K. Yang, O. Deng, C. Chen, R. Shin, S. Roy, and B. Van Durme, “Addressing Resource and Privacy Constraints in Semantic Parsing Through Data Augmentation,” in Findings of the Association for Computational Linguistics: ACL 2022, Dublin, Ireland, 2022, p. 3685–3695. doi:10.18653/v1/2022.findings-acl.291
[BibTeX] [Abstract] [Link]
We introduce a novel setup for low-resource task-oriented semantic parsing which incorporates several constraints that may arise in real-world scenarios: (1) lack of similar datasets/models from a related domain, (2) inability to sample useful logical forms directly from a grammar, and (3) privacy requirements for unlabeled natural utterances. Our goal is to improve a low-resource semantic parser using utterances collected through user interactions. In this highly challenging but realistic setting, we investigate data augmentation approaches involving generating a set of structured canonical utterances corresponding to logical forms, before simulating corresponding natural language and filtering the resulting pairs. We find that such approaches are effective despite our restrictive setup: in a low-resource setting on the complex SMCalFlow calendaring dataset (Andreas et al. 2020), we observe 33{\%} relative improvement over a non-data-augmented baseline in top-1 match.
@inproceedings{yang-etal-2022-addressing, title = "Addressing Resource and Privacy Constraints in Semantic Parsing Through Data Augmentation", author = "Yang, Kevin and Deng, Olivia and Chen, Charles and Shin, Richard and Roy, Subhro and Van Durme, Benjamin", editor = "Muresan, Smaranda and Nakov, Preslav and Villavicencio, Aline", booktitle = "Findings of the Association for Computational Linguistics: ACL 2022", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.findings-acl.291", doi = "10.18653/v1/2022.findings-acl.291", pages = "3685--3695", abstract = "We introduce a novel setup for low-resource task-oriented semantic parsing which incorporates several constraints that may arise in real-world scenarios: (1) lack of similar datasets/models from a related domain, (2) inability to sample useful logical forms directly from a grammar, and (3) privacy requirements for unlabeled natural utterances. Our goal is to improve a low-resource semantic parser using utterances collected through user interactions. In this highly challenging but realistic setting, we investigate data augmentation approaches involving generating a set of structured canonical utterances corresponding to logical forms, before simulating corresponding natural language and filtering the resulting pairs. We find that such approaches are effective despite our restrictive setup: in a low-resource setting on the complex SMCalFlow calendaring dataset (Andreas et al. 2020), we observe 33{\%} relative improvement over a non-data-augmented baseline in top-1 match.", } - S. Panthaplackel, A. Benton, and M. Dredze, “Updated Headline Generation: Creating Updated Summaries for Evolving News Stories,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Dublin, Ireland, 2022, p. 6438–6461. doi:10.18653/v1/2022.acl-long.446
[BibTeX] [Abstract] [Link]
We propose the task of updated headline generation, in which a system generates a headline for an updated article, considering both the previous article and headline. The system must identify the novel information in the article update, and modify the existing headline accordingly. We create data for this task using the NewsEdits corpus by automatically identifying contiguous article versions that are likely to require a substantive headline update. We find that models conditioned on the prior headline and body revisions produce headlines judged by humans to be as factual as gold headlines while making fewer unnecessary edits compared to a standard headline generation model. Our experiments establish benchmarks for this new contextual summarization task.
@inproceedings{panthaplackel-etal-2022-updated, title = "Updated Headline Generation: Creating Updated Summaries for Evolving News Stories", author = "Panthaplackel, Sheena and Benton, Adrian and Dredze, Mark", editor = "Muresan, Smaranda and Nakov, Preslav and Villavicencio, Aline", booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.acl-long.446", doi = "10.18653/v1/2022.acl-long.446", pages = "6438--6461", abstract = "We propose the task of updated headline generation, in which a system generates a headline for an updated article, considering both the previous article and headline. The system must identify the novel information in the article update, and modify the existing headline accordingly. We create data for this task using the NewsEdits corpus by automatically identifying contiguous article versions that are likely to require a substantive headline update. We find that models conditioned on the prior headline and body revisions produce headlines judged by humans to be as factual as gold headlines while making fewer unnecessary edits compared to a standard headline generation model. Our experiments establish benchmarks for this new contextual summarization task.", } - A. Anastasopoulos, L. Barrault, L. Bentivogli, M. Zanon Boito, O. Bojar, R. Cattoni, A. Currey, G. Dinu, K. Duh, M. Elbayad, C. Emmanuel, Y. Estève, M. Federico, C. Federmann, S. Gahbiche, H. Gong, R. Grundkiewicz, B. Haddow, B. Hsu, D. Javorský, V. Kloudová, S. Lakew, X. Ma, P. Mathur, P. McNamee, K. Murray, M. N{v{a}}dejde, S. Nakamura, M. Negri, J. Niehues, X. Niu, J. Ortega, J. Pino, E. Salesky, J. Shi, M. Sperber, S. Stüker, K. Sudoh, M. Turchi, Y. Virkar, A. Waibel, C. Wang, and S. Watanabe, “Findings of the IWSLT 2022 Evaluation Campaign,” in Proceedings of the 19th International Conference on Spoken Language Translation (IWSLT 2022), Dublin, Ireland (in-person and online), 2022, p. 98–157. doi:10.18653/v1/2022.iwslt-1.10
[BibTeX] [Abstract] [Link]
The evaluation campaign of the 19th International Conference on Spoken Language Translation featured eight shared tasks: (i) Simultaneous speech translation, (ii) Offline speech translation, (iii) Speech to speech translation, (iv) Low-resource speech translation, (v) Multilingual speech translation, (vi) Dialect speech translation, (vii) Formality control for speech translation, (viii) Isometric speech translation. A total of 27 teams participated in at least one of the shared tasks. This paper details, for each shared task, the purpose of the task, the data that were released, the evaluation metrics that were applied, the submissions that were received and the results that were achieved.
@inproceedings{anastasopoulos-etal-2022-findings, title = "Findings of the {IWSLT} 2022 Evaluation Campaign", author = {Anastasopoulos, Antonios and Barrault, Lo{\"\i}c and Bentivogli, Luisa and Zanon Boito, Marcely and Bojar, Ond{\v{r}}ej and Cattoni, Roldano and Currey, Anna and Dinu, Georgiana and Duh, Kevin and Elbayad, Maha and Emmanuel, Clara and Est{\`e}ve, Yannick and Federico, Marcello and Federmann, Christian and Gahbiche, Souhir and Gong, Hongyu and Grundkiewicz, Roman and Haddow, Barry and Hsu, Benjamin and Javorsk{\'y}, D{\'a}vid and Kloudov{\'a}, V{\u{e}}ra and Lakew, Surafel and Ma, Xutai and Mathur, Prashant and McNamee, Paul and Murray, Kenton and N{\v{a}}dejde, Maria and Nakamura, Satoshi and Negri, Matteo and Niehues, Jan and Niu, Xing and Ortega, John and Pino, Juan and Salesky, Elizabeth and Shi, Jiatong and Sperber, Matthias and St{\"u}ker, Sebastian and Sudoh, Katsuhito and Turchi, Marco and Virkar, Yogesh and Waibel, Alexander and Wang, Changhan and Watanabe, Shinji}, editor = "Salesky, Elizabeth and Federico, Marcello and Costa-juss{\`a}, Marta", booktitle = "Proceedings of the 19th International Conference on Spoken Language Translation (IWSLT 2022)", month = may, year = "2022", address = "Dublin, Ireland (in-person and online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.iwslt-1.10", doi = "10.18653/v1/2022.iwslt-1.10", pages = "98--157", abstract = "The evaluation campaign of the 19th International Conference on Spoken Language Translation featured eight shared tasks: (i) Simultaneous speech translation, (ii) Offline speech translation, (iii) Speech to speech translation, (iv) Low-resource speech translation, (v) Multilingual speech translation, (vi) Dialect speech translation, (vii) Formality control for speech translation, (viii) Isometric speech translation. A total of 27 teams participated in at least one of the shared tasks. This paper details, for each shared task, the purpose of the task, the data that were released, the evaluation metrics that were applied, the submissions that were received and the results that were achieved.", } - M. Yuan, P. Xia, C. May, B. Van Durme, and J. Boyd-Graber, “Adapting Coreference Resolution Models through Active Learning,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Dublin, Ireland, 2022, p. 7533–7549. doi:10.18653/v1/2022.acl-long.519
[BibTeX] [Abstract] [Link]
Neural coreference resolution models trained on one dataset may not transfer to new, low-resource domains. Active learning mitigates this problem by sampling a small subset of data for annotators to label. While active learning is well-defined for classification tasks, its application to coreference resolution is neither well-defined nor fully understood. This paper explores how to actively label coreference, examining sources of model uncertainty and document reading costs. We compare uncertainty sampling strategies and their advantages through thorough error analysis. In both synthetic and human experiments, labeling spans within the same document is more effective than annotating spans across documents. The findings contribute to a more realistic development of coreference resolution models.
@inproceedings{yuan-etal-2022-adapting, title = "Adapting Coreference Resolution Models through Active Learning", author = "Yuan, Michelle and Xia, Patrick and May, Chandler and Van Durme, Benjamin and Boyd-Graber, Jordan", editor = "Muresan, Smaranda and Nakov, Preslav and Villavicencio, Aline", booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.acl-long.519", doi = "10.18653/v1/2022.acl-long.519", pages = "7533--7549", abstract = "Neural coreference resolution models trained on one dataset may not transfer to new, low-resource domains. Active learning mitigates this problem by sampling a small subset of data for annotators to label. While active learning is well-defined for classification tasks, its application to coreference resolution is neither well-defined nor fully understood. This paper explores how to actively label coreference, examining sources of model uncertainty and document reading costs. We compare uncertainty sampling strategies and their advantages through thorough error analysis. In both synthetic and human experiments, labeling spans within the same document is more effective than annotating spans across documents. The findings contribute to a more realistic development of coreference resolution models.", } - S. Sun, A. Fan, J. Cross, V. Chaudhary, C. Tran, P. Koehn, and F. Guzmán, “Alternative Input Signals Ease Transfer in Multilingual Machine Translation,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Dublin, Ireland, 2022, p. 5291–5305. doi:10.18653/v1/2022.acl-long.363
[BibTeX] [Abstract] [Link]
Recent work in multilingual machine translation (MMT) has focused on the potential of positive transfer between languages, particularly cases where higher-resourced languages can benefit lower-resourced ones. While training an MMT model, the supervision signals learned from one language pair can be transferred to the other via the tokens shared by multiple source languages. However, the transfer is inhibited when the token overlap among source languages is small, which manifests naturally when languages use different writing systems. In this paper, we tackle inhibited transfer by augmenting the training data with alternative signals that unify different writing systems, such as phonetic, romanized, and transliterated input. We test these signals on Indic and Turkic languages, two language families where the writing systems differ but languages still share common features. Our results indicate that a straightforward multi-source self-ensemble {–} training a model on a mixture of various signals and ensembling the outputs of the same model fed with different signals during inference, outperforms strong ensemble baselines by 1.3 BLEU points on both language families. Further, we find that incorporating alternative inputs via self-ensemble can be particularly effective when training set is small, leading to +5 BLEU when only 5{\%} of the total training data is accessible. Finally, our analysis demonstrates that including alternative signals yields more consistency and translates named entities more accurately, which is crucial for increased factuality of automated systems.
@inproceedings{sun-etal-2022-alternative, title = "Alternative Input Signals Ease Transfer in Multilingual Machine Translation", author = "Sun, Simeng and Fan, Angela and Cross, James and Chaudhary, Vishrav and Tran, Chau and Koehn, Philipp and Guzm{\'a}n, Francisco", editor = "Muresan, Smaranda and Nakov, Preslav and Villavicencio, Aline", booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = may, year = "2022", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.acl-long.363", doi = "10.18653/v1/2022.acl-long.363", pages = "5291--5305", abstract = "Recent work in multilingual machine translation (MMT) has focused on the potential of positive transfer between languages, particularly cases where higher-resourced languages can benefit lower-resourced ones. While training an MMT model, the supervision signals learned from one language pair can be transferred to the other via the tokens shared by multiple source languages. However, the transfer is inhibited when the token overlap among source languages is small, which manifests naturally when languages use different writing systems. In this paper, we tackle inhibited transfer by augmenting the training data with alternative signals that unify different writing systems, such as phonetic, romanized, and transliterated input. We test these signals on Indic and Turkic languages, two language families where the writing systems differ but languages still share common features. Our results indicate that a straightforward multi-source self-ensemble {--} training a model on a mixture of various signals and ensembling the outputs of the same model fed with different signals during inference, outperforms strong ensemble baselines by 1.3 BLEU points on both language families. Further, we find that incorporating alternative inputs via self-ensemble can be particularly effective when training set is small, leading to +5 BLEU when only 5{\%} of the total training data is accessible. Finally, our analysis demonstrates that including alternative signals yields more consistency and translates named entities more accurately, which is crucial for increased factuality of automated systems.", } - J. Zhou, J. Eisner, M. Newman, E. A. Platanios, and S. Thomson, “Online Semantic Parsing for Latency Reduction in Task-Oriented Dialogue,” in Proceedings of the Association for Computational Linguistics (ACL), Dublin, 2022, p. 1554–1576. doi:10.18653/v1/2022.acl-long.110
[BibTeX] [Link]@InProceedings{zhou-et-al-2022, aclid = "2022.acl-long.110", doi = "10.18653/v1/2022.acl-long.110", author = "Jiawei Zhou and Jason Eisner and Michael Newman and Emmanouil Anthony Platanios and Sam Thomson", title = "Online Semantic Parsing for Latency Reduction in Task-Oriented Dialogue", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", pages = "1554--1576", year = "2022", month = may, address = "Dublin", URL = "http://cs.jhu.edu/~jason/papers/#zhou-et-al-2022", } - R. Cotterell and J. Eisner, “A Functionalist Account of Vowel System Typology,” in Proceedings of the Association for Computational Linguistics (ACL), Dublin, 2022.
[BibTeX] [Link]@InProceedings{cotterell-eisner-2022, author = "Ryan Cotterell and Jason Eisner", title = "A Functionalist Account of Vowel System Typology", booktitle = "Proceedings of the Association for Computational Linguistics (ACL)", year = "2022", month = may, address = "Dublin", URL = "http://cs.jhu.edu/~jason/papers/#cotterell-eisner-2022", } - C. Yang, H. Mei, and J. Eisner, “Transformer Embeddings of Irregularly Spaced Events and Their Participants,” in Proceedings of the Tenth International Conference on Learning Representations (ICLR), 2022.
[BibTeX] [Link]@InProceedings{yang-et-al-2022-iclr, author = "Chenghao Yang and Hongyuan Mei and Jason Eisner", title = "Transformer Embeddings of Irregularly Spaced Events and Their Participants", booktitle = "Proceedings of the Tenth International Conference on Learning Representations (ICLR)", year = "2022", month = apr, note = "9 pages plus appendices", URL = "http://cs.jhu.edu/~jason/papers/#yang-et-al-2022-iclr", } - I. Lin, L. Njoo, A. Field, A. Sharma, K. Reinecke, T. Althoff, and Y. Tsvetkov, “Gendered Mental Health Stigma in Masked Language Models,” in Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Abu Dhabi, United Arab Emirates, 2022, p. 2152–2170. doi:10.18653/v1/2022.emnlp-main.139
[BibTeX] [Abstract] [Link]
Mental health stigma prevents many individuals from receiving the appropriate care, and social psychology studies have shown that mental health tends to be overlooked in men. In this work, we investigate gendered mental health stigma in masked language models. In doing so, we operationalize mental health stigma by developing a framework grounded in psychology research: we use clinical psychology literature to curate prompts, then evaluate the models{‘} propensity to generate gendered words. We find that masked language models capture societal stigma about gender in mental health: models are consistently more likely to predict female subjects than male in sentences about having a mental health condition (32{\%} vs. 19{\%}), and this disparity is exacerbated for sentences that indicate treatment-seeking behavior. Furthermore, we find that different models capture dimensions of stigma differently for men and women, associating stereotypes like anger, blame, and pity more with women with mental health conditions than with men. In showing the complex nuances of models{‘} gendered mental health stigma, we demonstrate that context and overlapping dimensions of identity are important considerations when assessing computational models{‘} social biases.
@inproceedings{lin-etal-2022-gendered, title = "Gendered Mental Health Stigma in Masked Language Models", author = "Lin, Inna and Njoo, Lucille and Field, Anjalie and Sharma, Ashish and Reinecke, Katharina and Althoff, Tim and Tsvetkov, Yulia", editor = "Goldberg, Yoav and Kozareva, Zornitsa and Zhang, Yue", booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2022", address = "Abu Dhabi, United Arab Emirates", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.emnlp-main.139", doi = "10.18653/v1/2022.emnlp-main.139", pages = "2152--2170", abstract = "Mental health stigma prevents many individuals from receiving the appropriate care, and social psychology studies have shown that mental health tends to be overlooked in men. In this work, we investigate gendered mental health stigma in masked language models. In doing so, we operationalize mental health stigma by developing a framework grounded in psychology research: we use clinical psychology literature to curate prompts, then evaluate the models{'} propensity to generate gendered words. We find that masked language models capture societal stigma about gender in mental health: models are consistently more likely to predict female subjects than male in sentences about having a mental health condition (32{\%} vs. 19{\%}), and this disparity is exacerbated for sentences that indicate treatment-seeking behavior. Furthermore, we find that different models capture dimensions of stigma differently for men and women, associating stereotypes like anger, blame, and pity more with women with mental health conditions than with men. In showing the complex nuances of models{'} gendered mental health stigma, we demonstrate that context and overlapping dimensions of identity are important considerations when assessing computational models{'} social biases.", } - W. G. C. Bandara and Vishal M. Patel, “A Transformer-Based Siamese Network for Change Detection,” in IEEE International Geoscience and Remote Sensing Symposium, 2022.
[BibTeX] [Link]@inproceedings{245668909, title = {A Transformer-Based Siamese Network for Change Detection}, author = {{W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {1}, booktitle = {IEEE International Geoscience and Remote Sensing Symposium}, url = {https://www.semanticscholar.org/paper/ef3b15260a610473c95662f5df2c195ac19f64d6}, } - Tae Soo Kim, Bohoon Shim, Michael Peven, Weichao Qiu, A. Yuille, and Gregory Hager, “Learning from Synthetic Vehicles,” in 2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW), 2022.
[BibTeX] [Link]@inproceedings{246872267, title = {Learning from Synthetic Vehicles}, author = {{Tae Soo Kim} and {Bohoon Shim} and {Michael Peven} and {Weichao Qiu} and {A. Yuille} and {Gregory Hager}}, year = 2022, month = {1}, booktitle = {2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)}, url = {https://www.semanticscholar.org/paper/b5ac4931672397df6d9135e0d5b615351f490a44}, } - R. Yasarla, C. Priebe, and Vishal M. Patel, “ART-SS: An Adaptive Rejection Technique for Semi-Supervised restoration for adverse weather-affected images,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{247518863, title = {ART-SS: An Adaptive Rejection Technique for Semi-Supervised restoration for adverse weather-affected images}, author = {{R. Yasarla} and {C. Priebe} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/0bf4fd83f0f17b0fa94c18631a28d52ce5ea6042}, } - Tasnim Mohiuddin, Philipp Koehn, Vishrav Chaudhary, James Cross, Shruti Bhosale, and Shafiq R. Joty, “Data Selection Curriculum for Neural Machine Translation,” in Conference on Empirical Methods in Natural Language Processing, 2022.
[BibTeX] [Link]@inproceedings{247762191, title = {Data Selection Curriculum for Neural Machine Translation}, author = {{Tasnim Mohiuddin} and {Philipp Koehn} and {Vishrav Chaudhary} and {James Cross} and {Shruti Bhosale} and {Shafiq R. Joty}}, year = 2022, month = {3}, booktitle = {Conference on Empirical Methods in Natural Language Processing}, url = {https://www.semanticscholar.org/paper/d6c4b31958fe9e4ff4f83e049ed5c6881653eb03}, } - Kangfu Mei, Yiqun Mei, and Vishal M. Patel, “Thermal to Visible Image Synthesis Under Atmospheric Turbulence,” in International Conference on Information Photonics, 2022.
[BibTeX] [Link]@inproceedings{248006370, title = {Thermal to Visible Image Synthesis Under Atmospheric Turbulence}, author = {{Kangfu Mei} and {Yiqun Mei} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/0a123eb1a768cc151ff9ebb004cc2461414a53a3}, } - R. Arora, Raef Bassily, Tom’as Gonz’alez, Crist’obal Guzm’an, Michael Menart, and Enayat Ullah, “Faster Rates of Convergence to Stationary Points in Differentially Private Optimization,” in International Conference on Machine Learning, 2022.
[BibTeX] [Link]@inproceedings{249282662, title = {Faster Rates of Convergence to Stationary Points in Differentially Private Optimization}, author = {{R. Arora} and {Raef Bassily} and {Tom'as Gonz'alez} and {Crist'obal Guzm'an} and {Michael Menart} and {Enayat Ullah}}, year = 2022, month = {6}, booktitle = {International Conference on Machine Learning}, url = {https://www.semanticscholar.org/paper/6f85ad4e04fc157ed5b499e348972f188a39cd10}, } - Nithin Gopalakrishnan Nair, W. G. C. Bandara, and Vishal M. Patel, “Image Generation with Multimodal Priors using Denoising Diffusion Probabilistic Models,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249605363, title = {Image Generation with Multimodal Priors using Denoising Diffusion Probabilistic Models}, author = {{Nithin Gopalakrishnan Nair} and {W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/c6480d46777da8f0e5fa6e65760f0adec31e4bff}, } - Angtian Wang, Peng Wang, Jian Sun, Adam Kortylewski, and A. Yuille, “VoGE: A Differentiable Volume Renderer using Gaussian Ellipsoids for Analysis-by-Synthesis,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{249209554, title = {VoGE: A Differentiable Volume Renderer using Gaussian Ellipsoids for Analysis-by-Synthesis}, author = {{Angtian Wang} and {Peng Wang} and {Jian Sun} and {Adam Kortylewski} and {A. Yuille}}, year = 2022, month = {5}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/31e79b62a9483dcdf2575603469e6ff888e7f234}, } - A. Hussein, S. A. Chowdhury, Ahmed Abdelali, N. Dehak, and Ahmed M. Ali, “Code-Switching Text Augmentation for Multilingual Speech Processing,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{245827791, title = {Code-Switching Text Augmentation for Multilingual Speech Processing}, author = {{A. Hussein} and {S. A. Chowdhury} and {Ahmed Abdelali} and {N. Dehak} and {Ahmed M. Ali}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/be5074a85ef8166fc173cb51971a2e3f79685134}, } - Qihao Liu, Yi Zhang, S. Bai, and A. Yuille, “Explicit Occlusion Reasoning for Multi-person 3D Human Pose Estimation,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{251223772, title = {Explicit Occlusion Reasoning for Multi-person 3D Human Pose Estimation}, author = {{Qihao Liu} and {Yi Zhang} and {S. Bai} and {A. Yuille}}, year = 2022, month = {7}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/6daa8592e9c895acd78ce0d798f5327add2902a4}, } - Nupoor Gandhi, Anjalie Field, and Emma Strubell, “Mention Annotations Alone Enable Efficient Domain Adaptation for Coreference Resolution,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{260444131, title = {Mention Annotations Alone Enable Efficient Domain Adaptation for Coreference Resolution}, author = {{Nupoor Gandhi} and {Anjalie Field} and {Emma Strubell}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/43f09be116b87046334d395a71919ab423b204a1}, } - Jonah P. Sengupta, M. Villemur, P. Pouliquen, P. Julián, and A. Andreou, “Embedded Processing Pipeline Exploration For Neuromorphic Event Based Perceptual Systems,” in International Symposium on Circuits and Systems, 2022.
[BibTeX] [Link]@inproceedings{253461961, title = {Embedded Processing Pipeline Exploration For Neuromorphic Event Based Perceptual Systems}, author = {{Jonah P. Sengupta} and {M. Villemur} and {P. Pouliquen} and {P. Julián} and {A. Andreou}}, year = 2022, month = {5}, booktitle = {International Symposium on Circuits and Systems}, url = {https://www.semanticscholar.org/paper/42845a69a8efd8e8dc7b697c3ce0a4a8f6dfae86}, } - Yiwen Shao, J. Villalba, Sonal Joshi, Saurabh Kataria, S. Khudanpur, and N. Dehak, “Chunking Defense for Adversarial Attacks on ASR,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{252341100, title = {Chunking Defense for Adversarial Attacks on ASR}, author = {{Yiwen Shao} and {J. Villalba} and {Sonal Joshi} and {Saurabh Kataria} and {S. Khudanpur} and {N. Dehak}}, year = 2022, month = {9}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/ace27d0f6e93765439e19203e69570cf00f09e63}, } - Sangwook Park and Mounya Elhilali, “Time-Balanced Focal Loss for Audio Event Detection,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{249437208, title = {Time-Balanced Focal Loss for Audio Event Detection}, author = {{Sangwook Park} and {Mounya Elhilali}}, year = 2022, month = {5}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/62b7aa0300a9ebc3d494629579a4a051874b82a8}, } - H. Goldman, Julia Porcino, G. Divita, Ayah Zirikly, Bart Desmet, Maryanne Sacco, E. Marfeo, Christine M. McDonough, E. Rasch, and L. Chan, “Informatics Research on Mental Health Functioning: Decision Support for the Social Security Administration Disability Program.,” in Psychiatric Services, 2022.
[BibTeX] [Link]@inproceedings{249276757, title = {Informatics Research on Mental Health Functioning: Decision Support for the Social Security Administration Disability Program.}, author = {{H. Goldman} and {Julia Porcino} and {G. Divita} and {Ayah Zirikly} and {Bart Desmet} and {Maryanne Sacco} and {E. Marfeo} and {Christine M. McDonough} and {E. Rasch} and {L. Chan}}, year = 2022, month = {6}, booktitle = {Psychiatric Services}, url = {https://www.semanticscholar.org/paper/99410edf5a03b98ff66fa16e86bc39412fefa2e6}, } - Alex Mallen, Akari Asai, Victor Zhong, Rajarshi Das, Hannaneh Hajishirzi, and Daniel Khashabi, “When Not to Trust Language Models: Investigating Effectiveness and Limitations of Parametric and Non-Parametric Memories,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{260443047, title = {When Not to Trust Language Models: Investigating Effectiveness and Limitations of Parametric and Non-Parametric Memories}, author = {{Alex Mallen} and {Akari Asai} and {Victor Zhong} and {Rajarshi Das} and {Hannaneh Hajishirzi} and {Daniel Khashabi}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7b0f98f51040700aae3cd9f0e3432dedcd69fb30}, } - Y. Xia, Q. Yu, L. Chu, S. Kawamoto, S. Park, F. Liu, J. Chen, Z. Zhu, B. Li, Z. Zhou, Y. Lu, Y. Wang, W. Shen, L. Xie, Y. Zhou, elliot k fishman, A. Javed, D. Fouladi, S. Shayesteh, J. Graves, A. Blanco, E. Zinreich, B. Kinny-Koster, K. Kinzler, R. Hruban, B. Vogelstein, A. Yuille, and E. Fishman, “The FELIX Project: Deep Networks To Detect Pancreatic Neoplasms,” in medRxiv, 2022.
[BibTeX] [Link]@inproceedings{252519938, title = {The FELIX Project: Deep Networks To Detect Pancreatic Neoplasms}, author = {{Y. Xia} and {Q. Yu} and {L. Chu} and {S. Kawamoto} and {S. Park} and {F. Liu} and {J. Chen} and {Z. Zhu} and {B. Li} and {Z. Zhou} and {Y. Lu} and {Y. Wang} and {W. Shen} and {L. Xie} and {Y. Zhou} and {elliot k fishman} and {A. Javed} and {D. Fouladi} and {S. Shayesteh} and {J. Graves} and {A. Blanco} and {E. Zinreich} and {B. Kinny-Koster} and {K. Kinzler} and {R. Hruban} and {B. Vogelstein} and {A. Yuille} and {E. Fishman}}, year = 2022, month = {9}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/3167cedfe031711fa832f5ba48519357923ac0c7}, } - Anjalie Field, Chan Young Park, Antônio Theóphilo, J. Watson-Daniels, and Yulia Tsvetkov, “An analysis of emotions and the prominence of positivity in #BlackLivesMatter tweets,” in Proceedings of the National Academy of Sciences of the United States of America, 2022.
[BibTeX] [Link]@inproceedings{251766132, title = {An analysis of emotions and the prominence of positivity in #BlackLivesMatter tweets}, author = {{Anjalie Field} and {Chan Young Park} and {Antônio Theóphilo} and {J. Watson-Daniels} and {Yulia Tsvetkov}}, year = 2022, month = {8}, booktitle = {Proceedings of the National Academy of Sciences of the United States of America}, url = {https://www.semanticscholar.org/paper/6dadf66d41b5c5bf4ce8f49fce38bc4f44889246}, } - Nithin Gopalakrishnan Nair, R. Yasarla, and Vishal M. Patel, “NBD-GAP: Non-Blind Image Deblurring without Clean Target Images,” in International Conference on Information Photonics, 2022.
[BibTeX] [Link]@inproceedings{252383306, title = {NBD-GAP: Non-Blind Image Deblurring without Clean Target Images}, author = {{Nithin Gopalakrishnan Nair} and {R. Yasarla} and {Vishal M. Patel}}, year = 2022, month = {9}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/28a43c5d52c421b1ccc24d15f39b2cdb82ed84de}, } - B. Vasey, M. Nagendran, Bruce Campbell, D. Clifton, Gary S. Collins, Spiros C. Denaxas, A. Denniston, L. Faes, B. Geerts, Mudathir Ibrahim, Xiaoxuan Liu, B. Mateen, P. Mathur, M. Mccradden, L. Morgan, Johan Ordish, Campbell Rogers, S. Saria, D. Ting, P. Watkinson, W. Weber, P. Wheatstone, and P. McCulloch, “Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI,” in British medical journal, 2022.
[BibTeX] [Link]@inproceedings{248864294, title = {Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI}, author = {{B. Vasey} and {M. Nagendran} and {Bruce Campbell} and {D. Clifton} and {Gary S. Collins} and {Spiros C. Denaxas} and {A. Denniston} and {L. Faes} and {B. Geerts} and {Mudathir Ibrahim} and {Xiaoxuan Liu} and {B. Mateen} and {P. Mathur} and {M. Mccradden} and {L. Morgan} and {Johan Ordish} and {Campbell Rogers} and {S. Saria} and {D. Ting} and {P. Watkinson} and {W. Weber} and {P. Wheatstone} and {P. McCulloch}}, year = 2022, month = {5}, booktitle = {British medical journal}, url = {https://www.semanticscholar.org/paper/3a8c344f67d5081ead5f7dd5ebf0f760d69fc01d}, } - Yunjuan Wang, Enayat Ullah, Poorya Mianjy, and R. Arora, “Adversarial Robustness is at Odds with Lazy Training,” in Neural Information Processing Systems, 2022.
[BibTeX] [Link]@inproceedings{250243820, title = {Adversarial Robustness is at Odds with Lazy Training}, author = {{Yunjuan Wang} and {Enayat Ullah} and {Poorya Mianjy} and {R. Arora}}, year = 2022, month = {6}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/e2100da66c556f6ce3fbe904696fb0cec2aca2bf}, } - Samik Sadhu and H. Hermansky, “Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{247628260, title = {Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech}, author = {{Samik Sadhu} and {H. Hermansky}}, year = 2022, month = {3}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/35a36559a133981c17759aa573afea646abe40f6}, } - Anastasia Razdaibiedina, A. Khetan, Zohar S. Karnin, Daniel Khashabi, Vishaal Kapoor, and V. Madan, “Representation Projection Invariance Mitigates Representation Collapse.” 2022.
[BibTeX] [Link]@inproceedings{258588228, title = {Representation Projection Invariance Mitigates Representation Collapse}, author = {{Anastasia Razdaibiedina} and {A. Khetan} and {Zohar S. Karnin} and {Daniel Khashabi} and {Vishaal Kapoor} and {V. Madan}}, year = 2022, month = {5}, booktitle = {}, url = {https://www.semanticscholar.org/paper/3746b0e7370784d5242dc9d3fc3fd3853a34409b}, } - Aimon Rahman, W. G. C. Bandara, Jeya Maria Jose Valanarasu, I. Hacihaliloglu, and Vishal M. Patel, “Orientation-guided Graph Convolutional Network for Bone Surface Segmentation,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2022.
[BibTeX] [Link]@inproceedings{249848080, title = {Orientation-guided Graph Convolutional Network for Bone Surface Segmentation}, author = {{Aimon Rahman} and {W. G. C. Bandara} and {Jeya Maria Jose Valanarasu} and {I. Hacihaliloglu} and {Vishal M. Patel}}, year = 2022, month = {6}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/bdcd82545a729552d83ed920bd117718c9f6948f}, } - Haoran Xu and Kenton Murray, “Por Qué Não Utiliser Alla Språk? Mixed Training with Gradient Optimization in Few-Shot Cross-Lingual Transfer,” in NAACL-HLT, 2022.
[BibTeX] [Link]@inproceedings{248476149, title = {Por Qué Não Utiliser Alla Språk? Mixed Training with Gradient Optimization in Few-Shot Cross-Lingual Transfer}, author = {{Haoran Xu} and {Kenton Murray}}, year = 2022, month = {4}, booktitle = {NAACL-HLT}, url = {https://www.semanticscholar.org/paper/811a5c79d8c0f6f5b57697e7be0e84b5f9a94ce8}, } - Sonal Joshi, Saurabh Kataria, Yiwen Shao, Piotr Żelasko, J. Villalba, S. Khudanpur, and N. Dehak, “Defense against Adversarial Attacks on Hybrid Speech Recognition using Joint Adversarial Fine-tuning with Denoiser,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248069341, title = {Defense against Adversarial Attacks on Hybrid Speech Recognition using Joint Adversarial Fine-tuning with Denoiser}, author = {{Sonal Joshi} and {Saurabh Kataria} and {Yiwen Shao} and {Piotr Żelasko} and {J. Villalba} and {S. Khudanpur} and {N. Dehak}}, year = 2022, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/49011d1b139bbb65fe273fd9e4b2197cee237385}, } - P. McNamee and K. Duh, “The Multilingual Microblog Translation Corpus: Improving and Evaluating Translation of User-Generated Text,” in Proceedings of the Thirteenth Language Resources and Evaluation Conference, Marseille, France, 2022, p. 910–918.
[BibTeX] [Abstract] [Link]
Translation of the noisy, informal language found in social media has been an understudied problem, with a principal factor being the limited availability of translation corpora in many languages. To address this need we have developed a new corpus containing over 200,000 translations of microblog posts that supports translation of thirteen languages into English. The languages are: Arabic, Chinese, Farsi, French, German, Hindi, Korean, Pashto, Portuguese, Russian, Spanish, Tagalog, and Urdu. We are releasing these data as the Multilingual Microblog Translation Corpus to support futher research in translation of informal language. We establish baselines using this new resource, and we further demonstrate the utility of the corpus by conducting experiments with fine-tuning to improve translation quality from a high performing neural machine translation (NMT) system. Fine-tuning provided substantial gains, ranging from +3.4 to +11.1 BLEU. On average, a relative gain of 21{\%} was observed, demonstrating the utility of the corpus.
@inproceedings{mcnamee-duh-2022-multilingual, title = "The Multilingual Microblog Translation Corpus: Improving and Evaluating Translation of User-Generated Text", author = "McNamee, Paul and Duh, Kevin", editor = "Calzolari, Nicoletta and B{\'e}chet, Fr{\'e}d{\'e}ric and Blache, Philippe and Choukri, Khalid and Cieri, Christopher and Declerck, Thierry and Goggi, Sara and Isahara, Hitoshi and Maegaard, Bente and Mariani, Joseph and Mazo, H{\'e}l{\`e}ne and Odijk, Jan and Piperidis, Stelios", booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference", month = jun, year = "2022", address = "Marseille, France", publisher = "European Language Resources Association", url = "https://aclanthology.org/2022.lrec-1.96", pages = "910--918", abstract = "Translation of the noisy, informal language found in social media has been an understudied problem, with a principal factor being the limited availability of translation corpora in many languages. To address this need we have developed a new corpus containing over 200,000 translations of microblog posts that supports translation of thirteen languages into English. The languages are: Arabic, Chinese, Farsi, French, German, Hindi, Korean, Pashto, Portuguese, Russian, Spanish, Tagalog, and Urdu. We are releasing these data as the Multilingual Microblog Translation Corpus to support futher research in translation of informal language. We establish baselines using this new resource, and we further demonstrate the utility of the corpus by conducting experiments with fine-tuning to improve translation quality from a high performing neural machine translation (NMT) system. Fine-tuning provided substantial gains, ranging from +3.4 to +11.1 BLEU. On average, a relative gain of 21{\%} was observed, demonstrating the utility of the corpus.", } - A. Javaid, Fawzi Zghyer, Chang H Kim, Erin M. Spaulding, N. Isakadze, Jie Ding, Daniel Kargillis, Yumin Gao, Faisal Rahman, Donald E. Brown, S. Saria, Seth S. Martin, C. Kramer, R. Blumenthal, and F. Marvel, “Medicine 2032: The future of cardiovascular disease prevention with machine learning and digital health technology,” in American Journal of Preventive Cardiology, 2022.
[BibTeX] [Link]@inproceedings{251936798, title = {Medicine 2032: The future of cardiovascular disease prevention with machine learning and digital health technology}, author = {{A. Javaid} and {Fawzi Zghyer} and {Chang H Kim} and {Erin M. Spaulding} and {N. Isakadze} and {Jie Ding} and {Daniel Kargillis} and {Yumin Gao} and {Faisal Rahman} and {Donald E. Brown} and {S. Saria} and {Seth S. Martin} and {C. Kramer} and {R. Blumenthal} and {F. Marvel}}, year = 2022, month = {8}, booktitle = {American Journal of Preventive Cardiology}, url = {https://www.semanticscholar.org/paper/fe2f3307cb21f446a2e1272a008b2938cfd3d402}, } - Qihang Yu, Huiyu Wang, Siyuan Qiao, Maxwell D. Collins, Yukun Zhu, Hartwig Adam, A. Yuille, and Liang-Chieh Chen, “Appendix for k -means Mask Transformer.” 2022.
[BibTeX] [Link]@inproceedings{253513495, title = {Appendix for k -means Mask Transformer}, author = {{Qihang Yu} and {Huiyu Wang} and {Siyuan Qiao} and {Maxwell D. Collins} and {Yukun Zhu} and {Hartwig Adam} and {A. Yuille} and {Liang-Chieh Chen}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/46bfaa37e8b95f6bff810e5563d67e3404e78225}, } - Jaejin Cho, R. Pappagari, Piotr Żelasko, L. Moro-Velázquez, J. Villalba, and N. Dehak, “Non-Contrastive Self-Supervised Learning of Utterance-Level Speech Representations,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{251468156, title = {Non-Contrastive Self-Supervised Learning of Utterance-Level Speech Representations}, author = {{Jaejin Cho} and {R. Pappagari} and {Piotr Żelasko} and {L. Moro-Velázquez} and {J. Villalba} and {N. Dehak}}, year = 2022, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/f3d7789c627d3e62d92c225a272e408f287c6317}, } - J. Sadeghi, Kevin Duh, G. Sugiyama, V. Patel, G. Coppa, and R. Barrera, “Pre-hospital caloric deficits in surgical patients.,” in Nutrition and Health, 2022.
[BibTeX] [Link]@inproceedings{208391943, title = {Pre-hospital caloric deficits in surgical patients.}, author = {{J. Sadeghi} and {Kevin Duh} and {G. Sugiyama} and {V. Patel} and {G. Coppa} and {R. Barrera}}, year = 2022, month = {7}, booktitle = {Nutrition and Health}, url = {https://www.semanticscholar.org/paper/cfee21939b8a016ed3d947607940dc9a0ccf8b0c}, } - Nathaniel Weir and Benjamin Van Durme, “Dynamic Generation of Interpretable Inference Rules in a Neuro-Symbolic Expert System,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{252355543, title = {Dynamic Generation of Interpretable Inference Rules in a Neuro-Symbolic Expert System}, author = {{Nathaniel Weir} and {Benjamin Van Durme}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/fa46f4ddd5c6e793a47c61db9c1ecde7ea1c82bc}, } - VS Vibashan, Poojan Oza, and Vishal M. Patel, “Towards Online Domain Adaptive Object Detection,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{248085083, title = {Towards Online Domain Adaptive Object Detection}, author = {{VS Vibashan} and {Poojan Oza} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/ae1a767e40ce43b3cdcc2440a91dfe4a77cad901}, } - Junfeng Wu, Qihao Liu, Yi Jiang, S. Bai, A. Yuille, and Xiang Bai, “In Defense of Online Models for Video Instance Segmentation,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{250918749, title = {In Defense of Online Models for Video Instance Segmentation}, author = {{Junfeng Wu} and {Qihao Liu} and {Yi Jiang} and {S. Bai} and {A. Yuille} and {Xiang Bai}}, year = 2022, month = {7}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/65edfa85e5e665d51540a2c7ae1bcb6381793f68}, } - Yizhong Wang, Swaroop Mishra, Pegah Alipoormolabashi, Yeganeh Kordi, Amirreza Mirzaei, Anjana Arunkumar, Arjun Ashok, Arut Selvan Dhanasekaran, Atharva Naik, David Stap, Eshaan Pathak, Giannis Karamanolakis, Haizhi Gary Lai, I. Purohit, Ishani Mondal, Jacob Anderson, Kirby Kuznia, Krima Doshi, Maitreya Patel, Kuntal Kumar Pal, M. Moradshahi, Mihir Parmar, Mirali Purohit, Neeraj Varshney, Phani Rohitha Kaza, Pulkit Verma, Ravsehaj Singh Puri, Rushang Karia, Shailaja Keyur Sampat, Savan Doshi, S. Mishra, Sujan Reddy, Sumanta Patro, Tanay Dixit, Xudong Shen, Chitta Baral, Yejin Choi, Hannaneh Hajishirzi, Noah A. Smith, and Daniel Khashabi, “Benchmarking Generalization via In-Context Instructions on 1, 600+ Language Tasks,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248227391, title = {Benchmarking Generalization via In-Context Instructions on 1, 600+ Language Tasks}, author = {{Yizhong Wang} and {Swaroop Mishra} and {Pegah Alipoormolabashi} and {Yeganeh Kordi} and {Amirreza Mirzaei} and {Anjana Arunkumar} and {Arjun Ashok} and {Arut Selvan Dhanasekaran} and {Atharva Naik} and {David Stap} and {Eshaan Pathak} and {Giannis Karamanolakis} and {Haizhi Gary Lai} and {I. Purohit} and {Ishani Mondal} and {Jacob Anderson} and {Kirby Kuznia} and {Krima Doshi} and {Maitreya Patel} and {Kuntal Kumar Pal} and {M. Moradshahi} and {Mihir Parmar} and {Mirali Purohit} and {Neeraj Varshney} and {Phani Rohitha Kaza} and {Pulkit Verma} and {Ravsehaj Singh Puri} and {Rushang Karia} and {Shailaja Keyur Sampat} and {Savan Doshi} and {S. Mishra} and {Sujan Reddy} and {Sumanta Patro} and {Tanay Dixit} and {Xudong Shen} and {Chitta Baral} and {Yejin Choi} and {Hannaneh Hajishirzi} and {Noah A. Smith} and {Daniel Khashabi}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/ec64e324ce1210fe5245dfd0fb5a92058732e5b9}, } - Qihang Yu, Huiyu Wang, Dahun Kim, Siyuan Qiao, Maxwell D. Collins, Yukun Zhu, Hartwig Adam, A. Yuille, and Liang-Chieh Chen, “CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{249890221, title = {CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation}, author = {{Qihang Yu} and {Huiyu Wang} and {Dahun Kim} and {Siyuan Qiao} and {Maxwell D. Collins} and {Yukun Zhu} and {Hartwig Adam} and {A. Yuille} and {Liang-Chieh Chen}}, year = 2022, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/31a9744bd5421b3fbbad2ab38ce33bb2f352c77a}, } - Jinghao Zhou, Chen Wei, Huiyu Wang, Wei Shen, Cihang Xie, A. Yuille, and Tao Kong, “Image BERT Pre-training with Online Tokenizer,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{251647228, title = {Image BERT Pre-training with Online Tokenizer}, author = {{Jinghao Zhou} and {Chen Wei} and {Huiyu Wang} and {Wei Shen} and {Cihang Xie} and {A. Yuille} and {Tao Kong}}, year = 2022, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/ff169d09a933756e8798021dbf9e24a0bbfd9b38}, } - Orion Weller, Aleem Khan, Nathaniel Weir, Dawn J Lawrie, and Benjamin Van Durme, “Defending Against Poisoning Attacks in Open-Domain Question Answering,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{254877071, title = {Defending Against Poisoning Attacks in Open-Domain Question Answering}, author = {{Orion Weller} and {Aleem Khan} and {Nathaniel Weir} and {Dawn J Lawrie} and {Benjamin Van Durme}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7e44002c4f78458987a90dc7a0408d60dd5cdb7c}, } - Nithin Gopalakrishnan Nair and Vishal M. Patel, “T2V-DDPM: Thermal to Visible Face Translation using Denoising Diffusion Probabilistic Models,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2022.
[BibTeX] [Link]@inproceedings{252368348, title = {T2V-DDPM: Thermal to Visible Face Translation using Denoising Diffusion Probabilistic Models}, author = {{Nithin Gopalakrishnan Nair} and {Vishal M. Patel}}, year = 2022, month = {9}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/fc49634e80ab31929799786a97b7ea63834bbdb1}, } - M. Villemur, Jonah P. Sengupta, P. Julián, and A. Andreou, “Morphological, Object Detection Framework for Embedded, Event-based Sensing,” in International Conference on Event-Based Control, Communication, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{251762249, title = {Morphological, Object Detection Framework for Embedded, Event-based Sensing}, author = {{M. Villemur} and {Jonah P. Sengupta} and {P. Julián} and {A. Andreou}}, year = 2022, month = {6}, booktitle = {International Conference on Event-Based Control, Communication, and Signal Processing}, url = {https://www.semanticscholar.org/paper/dc774c02c8260a15a0098b2a193b7b5db7e3fdb1}, } - Cheng Peng, Pengfei Guo, S. K. Zhou, Vishal M. Patel, and Ramalingam Chellappa, “Towards performant and reliable undersampled MR reconstruction via diffusion model sampling,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2022.
[BibTeX] [Link]@inproceedings{247318941, title = {Towards performant and reliable undersampled MR reconstruction via diffusion model sampling}, author = {{Cheng Peng} and {Pengfei Guo} and {S. K. Zhou} and {Vishal M. Patel} and {Ramalingam Chellappa}}, year = 2022, month = {3}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/6a0bc6e194608b9e41cbf69794343888edb54378}, } - VS Vibashan, Jeya Maria Jose Valanarasu, and Vishal M. Patel, “Target and Task specific Source-Free Domain Adaptive Image Segmentation,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{247778672, title = {Target and Task specific Source-Free Domain Adaptive Image Segmentation}, author = {{VS Vibashan} and {Jeya Maria Jose Valanarasu} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/db37fdfed1260f94ffb08a174e3e19f28dd8835e}, } - Qihang Yu, Huiyu Wang, Siyuan Qiao, Maxwell D. Collins, Yukun Zhu, Hatwig Adam, A. Yuille, and Liang-Chieh Chen, “k-means Mask Transformer,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{250408015, title = {k-means Mask Transformer}, author = {{Qihang Yu} and {Huiyu Wang} and {Siyuan Qiao} and {Maxwell D. Collins} and {Yukun Zhu} and {Hatwig Adam} and {A. Yuille} and {Liang-Chieh Chen}}, year = 2022, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/5deef7fc161cbdc884aff15b9810f8a432c1489a}, } - Gerardo Flores, George H. Chen, T. Pollard, Ayah Zirikly, Michael C. Hughes, Tasmie Sarker, Joyce Ho, and Tristan Naumann, “Conference on Health, Inference, and Learning (CHIL) 2022,” in ACM Conference on Health, Inference, and Learning, 2022.
[BibTeX] [Link]@inproceedings{261102271, title = {Conference on Health, Inference, and Learning (CHIL) 2022}, author = {{Gerardo Flores} and {George H. Chen} and {T. Pollard} and {Ayah Zirikly} and {Michael C. Hughes} and {Tasmie Sarker} and {Joyce Ho} and {Tristan Naumann}}, year = 2022, booktitle = {ACM Conference on Health, Inference, and Learning}, url = {https://www.semanticscholar.org/paper/20d7a0ea43dfc3c086fd41ca90f8885ea892f965}, } - Jeya Maria Jose Valanarasu, Pengfei Guo, VS Vibashan, and Vishal M. Patel, “On-the-Fly Test-time Adaptation for Medical Image Segmentation,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{247411201, title = {On-the-Fly Test-time Adaptation for Medical Image Segmentation}, author = {{Jeya Maria Jose Valanarasu} and {Pengfei Guo} and {VS Vibashan} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3b8c4a2a005df6dc7e9fb0b9e2e81a887ace5a6c}, } - Jian Xue, Peidong Wang, Jinyu Li, Matt Post, and Yashesh Gaur, “Large-Scale Streaming End-to-End Speech Translation with Neural Transducers,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{248118691, title = {Large-Scale Streaming End-to-End Speech Translation with Neural Transducers}, author = {{Jian Xue} and {Peidong Wang} and {Jinyu Li} and {Matt Post} and {Yashesh Gaur}}, year = 2022, month = {4}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/5a5704382fd8c980937e10618713d641c846b313}, } - Piotr Żelasko, Siyuan Feng, Laureano Moro Velázquez, A. Abavisani, Saurabhchand Bhati, O. Scharenborg, M. Hasegawa-Johnson, and N. Dehak, “Discovering Phonetic Inventories with Crosslingual Automatic Speech Recognition,” in Computer Speech and Language, 2022.
[BibTeX] [Link]@inproceedings{246294754, title = {Discovering Phonetic Inventories with Crosslingual Automatic Speech Recognition}, author = {{Piotr Żelasko} and {Siyuan Feng} and {Laureano Moro Velázquez} and {A. Abavisani} and {Saurabhchand Bhati} and {O. Scharenborg} and {M. Hasegawa-Johnson} and {N. Dehak}}, year = 2022, month = {1}, booktitle = {Computer Speech and Language}, url = {https://www.semanticscholar.org/paper/9da09ca7192a7546728575b2c0dfb923a36f110f}, } - Nathaniel Weir and Benjamin Van Durme, “Dynamic Generation of Grounded Logical Explanations in a Neuro-Symbolic Expert System.” 2022.
[BibTeX] [Link]@inproceedings{258762900, title = {Dynamic Generation of Grounded Logical Explanations in a Neuro-Symbolic Expert System}, author = {{Nathaniel Weir} and {Benjamin Van Durme}}, year = 2022, month = {9}, booktitle = {}, url = {https://www.semanticscholar.org/paper/767c1338c001fdadb5bf4eed530fc78ede49e9ae}, } - Nllb team, M. Costa-jussà, James Cross, Onur cCelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Alison Youngblood, Bapi Akula, Loïc Barrault, Gabriel Mejia Gonzalez, Prangthip Hansanti, John Hoffman, Semarley Jarrett, Kaushik Ram Sadagopan, Dirk Rowe, S. Spruit, C. Tran, Pierre Yves Andrews, N. F. Ayan, Shruti Bhosale, Sergey Edunov, Angela Fan, Cynthia Gao, Vedanuj Goswami, Francisco Guzm’an, Philipp Koehn, Alexandre Mourachko, C. Ropers, Safiyyah Saleem, Holger Schwenk, and Jeff Wang, “No Language Left Behind: Scaling Human-Centered Machine Translation,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{250425961, title = {No Language Left Behind: Scaling Human-Centered Machine Translation}, author = {{Nllb team} and {M. Costa-jussà} and {James Cross} and {Onur cCelebi} and {Maha Elbayad} and {Kenneth Heafield} and {Kevin Heffernan} and {Elahe Kalbassi} and {Janice Lam} and {Daniel Licht} and {Jean Maillard} and {Anna Sun} and {Skyler Wang} and {Guillaume Wenzek} and {Alison Youngblood} and {Bapi Akula} and {Loïc Barrault} and {Gabriel Mejia Gonzalez} and {Prangthip Hansanti} and {John Hoffman} and {Semarley Jarrett} and {Kaushik Ram Sadagopan} and {Dirk Rowe} and {S. Spruit} and {C. Tran} and {Pierre Yves Andrews} and {N. F. Ayan} and {Shruti Bhosale} and {Sergey Edunov} and {Angela Fan} and {Cynthia Gao} and {Vedanuj Goswami} and {Francisco Guzm'an} and {Philipp Koehn} and {Alexandre Mourachko} and {C. Ropers} and {Safiyyah Saleem} and {Holger Schwenk} and {Jeff Wang}}, year = 2022, month = {7}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/0e638ce20f3e9b4dd1c10c32a29495c798425e63}, } - R. Adams, K. Henry, A. Sridharan, Hossein Soleimani, A. Zhan, Nishi Rawat, Lauren Johnson, D. Hager, S. Cosgrove, Andrew Markowski, E. Klein, E. Chen, M. Saheed, Maureen Henley, S. Miranda, Katrina Houston, Robert C. Linton, Anushree R Ahluwalia, Albert W. Wu, and S. Saria, “Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis,” in Nature Network Boston, 2022.
[BibTeX] [Link]@inproceedings{250954558, title = {Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis}, author = {{R. Adams} and {K. Henry} and {A. Sridharan} and {Hossein Soleimani} and {A. Zhan} and {Nishi Rawat} and {Lauren Johnson} and {D. Hager} and {S. Cosgrove} and {Andrew Markowski} and {E. Klein} and {E. Chen} and {M. Saheed} and {Maureen Henley} and {S. Miranda} and {Katrina Houston} and {Robert C. Linton} and {Anushree R Ahluwalia} and {Albert W. Wu} and {S. Saria}}, year = 2022, month = {7}, booktitle = {Nature Network Boston}, url = {https://www.semanticscholar.org/paper/9ad55e7b87e1557983bdef0e9fe7eb0f4254dd94}, } - Xianhang Li, Huiyu Wang, Chen Wei, Jieru Mei, A. Yuille, Yuyin Zhou, and Cihang Xie, “In Defense of Image Pre-Training for Spatiotemporal Recognition,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{248512944, title = {In Defense of Image Pre-Training for Spatiotemporal Recognition}, author = {{Xianhang Li} and {Huiyu Wang} and {Chen Wei} and {Jieru Mei} and {A. Yuille} and {Yuyin Zhou} and {Cihang Xie}}, year = 2022, month = {5}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/c0c0139333b9c642fe7789f4fe8f27bc647c280d}, } - Xiangyu Zhang, Zhanhong He, Shuyu Li, R. Togneri, and Leibny Paola García-Perera, “Investigating self-supervised learning for lyrics recognition,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{252531266, title = {Investigating self-supervised learning for lyrics recognition}, author = {{Xiangyu Zhang} and {Zhanhong He} and {Shuyu Li} and {R. Togneri} and {Leibny Paola García-Perera}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/6632436fd0a465c7b1399c503396233eb9d88b0e}, } - M. Naphade, Shuo Wang, D. Anastasiu, Zheng Tang, Ming-Ching Chang, Xiaodong Yang, Liang Zheng, Anuj Sharma, R. Chellappa, and Pranamesh Chakraborty, “The 6th AI City Challenge,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2022.
[BibTeX] [Link]@inproceedings{216914509, title = {The 6th AI City Challenge}, author = {{M. Naphade} and {Shuo Wang} and {D. Anastasiu} and {Zheng Tang} and {Ming-Ching Chang} and {Xiaodong Yang} and {Liang Zheng} and {Anuj Sharma} and {R. Chellappa} and {Pranamesh Chakraborty}}, year = 2022, month = {4}, booktitle = {2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)}, url = {https://www.semanticscholar.org/paper/7f489232a16a54fa2b11d5758101f078f9db797c}, } - A. Hussein, S. A. Chowdhury, Ahmed Abdelali, N. Dehak, Ahmed M. Ali, and S. Khudanpur, “Textual Data Augmentation for Arabic-English Code-Switching Speech Recognition,” in Spoken Language Technology Workshop, 2022.
[BibTeX] [Link]@inproceedings{255595965, title = {Textual Data Augmentation for Arabic-English Code-Switching Speech Recognition}, author = {{A. Hussein} and {S. A. Chowdhury} and {Ahmed Abdelali} and {N. Dehak} and {Ahmed M. Ali} and {S. Khudanpur}}, year = 2022, month = {1}, booktitle = {Spoken Language Technology Workshop}, url = {https://www.semanticscholar.org/paper/3c00e6cc82b49f046b5f36e5d5f8aa4af68cad5a}, } - R. Yasarla, Vishwanath A. Sindagi, and Vishal M. Patel, “Unsupervised Restoration of Weather-affected Images using Deep Gaussian Process-based CycleGAN,” in International Conference on Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{248377080, title = {Unsupervised Restoration of Weather-affected Images using Deep Gaussian Process-based CycleGAN}, author = {{R. Yasarla} and {Vishwanath A. Sindagi} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {International Conference on Pattern Recognition}, url = {https://www.semanticscholar.org/paper/ee48b57139e1d84c60926796195f5f77c2d8b1db}, } - Sonal Joshi, Saurabh Kataria, J. Villalba, and N. Dehak, “AdvEst: Adversarial Perturbation Estimation to Classify and Detect Adversarial Attacks against Speaker Identification,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{248069457, title = {AdvEst: Adversarial Perturbation Estimation to Classify and Detect Adversarial Attacks against Speaker Identification}, author = {{Sonal Joshi} and {Saurabh Kataria} and {J. Villalba} and {N. Dehak}}, year = 2022, month = {4}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/a8144dbb8481cb78e08fc34e452603984bb5aa01}, } - Lianhui Qin, S. Welleck, Daniel Khashabi, and Yejin Choi, “COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics,” in Neural Information Processing Systems, 2022.
[BibTeX] [Link]@inproceedings{247058662, title = {COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics}, author = {{Lianhui Qin} and {S. Welleck} and {Daniel Khashabi} and {Yejin Choi}}, year = 2022, month = {2}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/4a6a65968a8eb8c09ffb57a7774ddabb596565b1}, } - Ramraj Chandradevan, Eugene Yang, M. Yarmohammadi, and Eugene Agichtein, “Learning to Enrich Query Representation with Pseudo-Relevance Feedback for Cross-lingual Retrieval,” in Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022.
[BibTeX] [Link]@inproceedings{250340438, title = {Learning to Enrich Query Representation with Pseudo-Relevance Feedback for Cross-lingual Retrieval}, author = {{Ramraj Chandradevan} and {Eugene Yang} and {M. Yarmohammadi} and {Eugene Agichtein}}, year = 2022, month = {7}, booktitle = {Annual International ACM SIGIR Conference on Research and Development in Information Retrieval}, url = {https://www.semanticscholar.org/paper/f0c4f3cb741548c70a4db105fee227fc4f59dfd2}, } - Shao-Yuan Lo and Vishal M. Patel, “Exploring Adversarially Robust Training for Unsupervised Domain Adaptation,” in Asian Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{246996539, title = {Exploring Adversarially Robust Training for Unsupervised Domain Adaptation}, author = {{Shao-Yuan Lo} and {Vishal M. Patel}}, year = 2022, month = {2}, booktitle = {Asian Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/1329a9e14f6454227dfb584a57a910ef168f6a7d}, } - Yutong Bai, Zeyu Wang, Junfei Xiao, Chen Wei, Huiyu Wang, A. Yuille, Yuyin Zhou, and Cihang Xie, “Masked Autoencoders Enable Efficient Knowledge Distillers,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{251800257, title = {Masked Autoencoders Enable Efficient Knowledge Distillers}, author = {{Yutong Bai} and {Zeyu Wang} and {Junfei Xiao} and {Chen Wei} and {Huiyu Wang} and {A. Yuille} and {Yuyin Zhou} and {Cihang Xie}}, year = 2022, month = {8}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/a7cd547c539d69f99f17855242cb07bd80047f9a}, } - Vipul Gupta, Zhuowan Li, Adam Kortylewski, Chenyu Zhang, Yingwei Li, and A. Yuille, “SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context in Visual Question Answering,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{247958394, title = {SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context in Visual Question Answering}, author = {{Vipul Gupta} and {Zhuowan Li} and {Adam Kortylewski} and {Chenyu Zhang} and {Yingwei Li} and {A. Yuille}}, year = 2022, month = {4}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/0d2f848fff121133b3b77c7e691c6a2ba502be47}, } - Xiaolei Huang, Franck Dernoncourt, and Mark Dredze, “Enriching Unsupervised User Embedding via Medical Concepts,” in ACM Conference on Health, Inference, and Learning, 2022.
[BibTeX] [Link]@inproceedings{247594586, title = {Enriching Unsupervised User Embedding via Medical Concepts}, author = {{Xiaolei Huang} and {Franck Dernoncourt} and {Mark Dredze}}, year = 2022, month = {3}, booktitle = {ACM Conference on Health, Inference, and Learning}, url = {https://www.semanticscholar.org/paper/78a4f90b348f5401e8fb6b84bca0e539142b2530}, } - VS Vibashan, Poojan Oza, and Vishal M. Patel, “Instance Relation Graph Guided Source-Free Domain Adaptive Object Detection,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{247778633, title = {Instance Relation Graph Guided Source-Free Domain Adaptive Object Detection}, author = {{VS Vibashan} and {Poojan Oza} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/c850d77f3ce8e8fa989cc4f7b466b63b113fd6db}, } - Malsha V. Perera, W. G. C. Bandara, Jeya Maria Jose Valanarasu, and Vishal M. Patel, “SAR Despeckling Using Overcomplete Convolutional Networks,” in IEEE International Geoscience and Remote Sensing Symposium, 2022.
[BibTeX] [Link]@inproceedings{249209486, title = {SAR Despeckling Using Overcomplete Convolutional Networks}, author = {{Malsha V. Perera} and {W. G. C. Bandara} and {Jeya Maria Jose Valanarasu} and {Vishal M. Patel}}, year = 2022, month = {5}, booktitle = {IEEE International Geoscience and Remote Sensing Symposium}, url = {https://www.semanticscholar.org/paper/c4911e20fb50f6da552c812bda9ef4fdb525b939}, } - Samik Sadhu and H. Hermansky, “Blind Signal Dereverberation for Machine Speech Recognition,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{252683281, title = {Blind Signal Dereverberation for Machine Speech Recognition}, author = {{Samik Sadhu} and {H. Hermansky}}, year = 2022, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d2fda509740eede59e46892958531088f3f25aed}, } - Christos Sapsanis, M. Villemur, and A. Andreou, “Real Number Modeling of a SAR ADC behavior using SystemVerilog,” in International Conference on Synthesis, Modeling, Analysis and Simulation Methods and Applications to Circuit Design, 2022.
[BibTeX] [Link]@inproceedings{250463643, title = {Real Number Modeling of a SAR ADC behavior using SystemVerilog}, author = {{Christos Sapsanis} and {M. Villemur} and {A. Andreou}}, year = 2022, month = {6}, booktitle = {International Conference on Synthesis, Modeling, Analysis and Simulation Methods and Applications to Circuit Design}, url = {https://www.semanticscholar.org/paper/528b50e00ed3efece80bbc4557ecf4f8df98094a}, } - Yingwei Li, A. Yu, Tianjian Meng, Benjamin Caine, Jiquan Ngiam, Daiyi Peng, Junyang Shen, Bo-Xun Wu, Yifeng Lu, Denny Zhou, Quoc V. Le, A. Yuille, and Mingxing Tan, “DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{247476162, title = {DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection}, author = {{Yingwei Li} and {A. Yu} and {Tianjian Meng} and {Benjamin Caine} and {Jiquan Ngiam} and {Daiyi Peng} and {Junyang Shen} and {Bo-Xun Wu} and {Yifeng Lu} and {Denny Zhou} and {Quoc V. Le} and {A. Yuille} and {Mingxing Tan}}, year = 2022, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/5ffca96f4becdab649f085699594caa7c5c03e86}, } - W. Wu and D. Yarowsky, “Known Words Will Do: Unknown Concept Translation via Lexical Relations,” in Proceedings of the Fifth Workshop on Technologies for Machine Translation of Low-Resource Languages (LoResMT 2022), Gyeongju, Republic of Korea, 2022, p. 15–22.
[BibTeX] [Abstract] [Link]
Translating into low-resource languages is challenging due to the scarcity of training data. In this paper, we propose a probabilistic lexical translation method that bridges through lexical relations including synonyms, hypernyms, hyponyms, and co-hyponyms. This method, which only requires a dictionary like Wiktionary and a lexical database like WordNet, enables the translation of unknown vocabulary into low-resource languages for which we may only know the translation of a related concept. Experiments on translating a core vocabulary set into 472 languages, most of them low-resource, show the effectiveness of our approach.
@inproceedings{wu-yarowsky-2022-known, title = "Known Words Will Do: Unknown Concept Translation via Lexical Relations", author = "Wu, Winston and Yarowsky, David", editor = "Ojha, Atul Kr. and Liu, Chao-Hong and Vylomova, Ekaterina and Abbott, Jade and Washington, Jonathan and Oco, Nathaniel and Pirinen, Tommi A and Malykh, Valentin and Logacheva, Varvara and Zhao, Xiaobing", booktitle = "Proceedings of the Fifth Workshop on Technologies for Machine Translation of Low-Resource Languages (LoResMT 2022)", month = oct, year = "2022", address = "Gyeongju, Republic of Korea", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.loresmt-1.3", pages = "15--22", abstract = "Translating into low-resource languages is challenging due to the scarcity of training data. In this paper, we propose a probabilistic lexical translation method that bridges through lexical relations including synonyms, hypernyms, hyponyms, and co-hyponyms. This method, which only requires a dictionary like Wiktionary and a lexical database like WordNet, enables the translation of unknown vocabulary into low-resource languages for which we may only know the translation of a related concept. Experiments on translating a core vocabulary set into 472 languages, most of them low-resource, show the effectiveness of our approach.", } - Rui Shao, Pramuditha Perera, P. Yuen, and Vishal M. Patel, “Open-Set Adversarial Defense with Clean-Adversarial Mutual Learning,” in International Journal of Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{246823296, title = {Open-Set Adversarial Defense with Clean-Adversarial Mutual Learning}, author = {{Rui Shao} and {Pramuditha Perera} and {P. Yuen} and {Vishal M. Patel}}, year = 2022, month = {2}, booktitle = {International Journal of Computer Vision}, url = {https://www.semanticscholar.org/paper/bce77cb22110eaf52438cf03b8668b875c699c46}, } - Samik Sadhu and H. Hermansky, “Importance of Different Temporal Modulations of Speech: a Tale of two Perspectives,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{247922750, title = {Importance of Different Temporal Modulations of Speech: a Tale of two Perspectives}, author = {{Samik Sadhu} and {H. Hermansky}}, year = 2022, month = {3}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/3ce501d4d81d9a78c2e506df7f6de0d79ca91a5b}, } - Ayah Zirikly, Bart Desmet, D. Newman-Griffis, E. Marfeo, C. McDonough, Howard Goldman, and L. Chan, “Information Extraction Framework for Disability Determination Using a Mental Functioning Use-Case,” in JMIR Medical Informatics, 2022.
[BibTeX] [Link]@inproceedings{247520370, title = {Information Extraction Framework for Disability Determination Using a Mental Functioning Use-Case}, author = {{Ayah Zirikly} and {Bart Desmet} and {D. Newman-Griffis} and {E. Marfeo} and {C. McDonough} and {Howard Goldman} and {L. Chan}}, year = 2022, month = {3}, booktitle = {JMIR Medical Informatics}, url = {https://www.semanticscholar.org/paper/66ce3e5f86256fb9b54ab94457b3aa6a0080e6b2}, } - Jeya Maria Jose Valanarasu, He Zhang, Jianming Zhang, Yilin Wang, Zhe Lin, J. Echevarria, Yinglan Ma, Zijun Wei, Kalyan Sunkavalli, and Vishal M. Patel, “Interactive Portrait Harmonization,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{247476364, title = {Interactive Portrait Harmonization}, author = {{Jeya Maria Jose Valanarasu} and {He Zhang} and {Jianming Zhang} and {Yilin Wang} and {Zhe Lin} and {J. Echevarria} and {Yinglan Ma} and {Zijun Wei} and {Kalyan Sunkavalli} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/c423c8ef2d8101676a4c2ba403ad5970c0364f09}, } - Mo Zhou and Vishal M. Patel, “Enhancing Adversarial Robustness for Deep Metric Learning,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{247223074, title = {Enhancing Adversarial Robustness for Deep Metric Learning}, author = {{Mo Zhou} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/5bcdc704df91b425b76fc6b64f1582667505cfae}, } - Weiting Tan and Philipp Koehn, “Bitext Mining for Low-Resource Languages via Contrastive Learning,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{251765426, title = {Bitext Mining for Low-Resource Languages via Contrastive Learning}, author = {{Weiting Tan} and {Philipp Koehn}}, year = 2022, month = {8}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/767853fdd964e043c485ebb92afdcdf3ee8457e8}, } - R. Arora, Raef Bassily, Crist’obal Guzm’an, Michael Menart, and Enayat Ullah, “Differentially Private Generalized Linear Models Revisited,” in Neural Information Processing Systems, 2022.
[BibTeX] [Link]@inproceedings{248562546, title = {Differentially Private Generalized Linear Models Revisited}, author = {{R. Arora} and {Raef Bassily} and {Crist'obal Guzm'an} and {Michael Menart} and {Enayat Ullah}}, year = 2022, month = {5}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/6c1ae5bd9aa32a08141bf630c90b00523ef4ffb1}, } - Chan Young Park, Julia Mendelsohn, Anjalie Field, and Yulia Tsvetkov, “Challenges and Opportunities in Information Manipulation Detection: An Examination of Wartime Russian Media,” in Conference on Empirical Methods in Natural Language Processing, 2022.
[BibTeX] [Link]@inproceedings{253107926, title = {Challenges and Opportunities in Information Manipulation Detection: An Examination of Wartime Russian Media}, author = {{Chan Young Park} and {Julia Mendelsohn} and {Anjalie Field} and {Yulia Tsvetkov}}, year = 2022, month = {5}, booktitle = {Conference on Empirical Methods in Natural Language Processing}, url = {https://www.semanticscholar.org/paper/b616154578751e156b21561e1a5d5ed833a3506f}, } - Saurabh Kataria, J. Villalba, Laureano Moro-Vel’azquez, and N. Dehak, “Joint domain adaptation and speech bandwidth extension using time-domain GANs for speaker verification,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{247839251, title = {Joint domain adaptation and speech bandwidth extension using time-domain GANs for speaker verification}, author = {{Saurabh Kataria} and {J. Villalba} and {Laureano Moro-Vel'azquez} and {N. Dehak}}, year = 2022, month = {3}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/d58ebbc34e8ea987da5dda1bb132823b3e9105d3}, } - Sandeep Reddy Kothinti and Mounya Elhilali, “Temporal Contrastive-Loss for Audio Event Detection,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{249437588, title = {Temporal Contrastive-Loss for Audio Event Detection}, author = {{Sandeep Reddy Kothinti} and {Mounya Elhilali}}, year = 2022, month = {5}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/34fe9aa0f5e26768d196087ed146e2b3a576d73e}, } - Sai Saketh Rambhatla, Saksham Suri, R. Chellappa, and Abhinav Shrivastava, “SparseDet: Improving Sparsely Annotated Object Detection with Pseudo-positive Mining.” 2022.
[BibTeX] [Link]@inproceedings{245877805, title = {SparseDet: Improving Sparsely Annotated Object Detection with Pseudo-positive Mining}, author = {{Sai Saketh Rambhatla} and {Saksham Suri} and {R. Chellappa} and {Abhinav Shrivastava}}, year = 2022, month = {1}, booktitle = {}, url = {https://www.semanticscholar.org/paper/7f71d5804fe434168643babc616a76eb65d5882e}, } - R. Adams, K. Henry, and S. Saria, “Addressing the ‘coin flip model’ and the role of ‘process of care’ variables in the analysis of TREWS,” in medRxiv, 2022.
[BibTeX] [Link]@inproceedings{252333098, title = {Addressing the 'coin flip model' and the role of 'process of care' variables in the analysis of TREWS}, author = {{R. Adams} and {K. Henry} and {S. Saria}}, year = 2022, month = {9}, booktitle = {medRxiv}, url = {https://www.semanticscholar.org/paper/39383cc7a62fdd63e05873096d7283d5f1b90d59}, } - Jieneng Chen, Shuyang Sun, Ju He, Philip H. S. Torr, A. Yuille, and Song Bai, “Supplementary Materials: ”TransMix: Attend to Mix for Vision Transformers”.” 2022.
[BibTeX] [Link]@inproceedings{249877931, title = {Supplementary Materials: ”TransMix: Attend to Mix for Vision Transformers”}, author = {{Jieneng Chen} and {Shuyang Sun} and {Ju He} and {Philip H. S. Torr} and {A. Yuille} and {Song Bai}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/d378dc21ab5cfbde24b295ab759c9947f820bc94}, } - Wufei Ma, Angtian Wang, A. Yuille, and Adam Kortylewski, “Robust Category-Level 6D Pose Estimation with Coarse-to-Fine Rendering of Neural Features,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{252211999, title = {Robust Category-Level 6D Pose Estimation with Coarse-to-Fine Rendering of Neural Features}, author = {{Wufei Ma} and {Angtian Wang} and {A. Yuille} and {Adam Kortylewski}}, year = 2022, month = {9}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/efa699cba13396c1b6d05a0dea9840020d29ae57}, } - Sangwook Park, Sandeep Reddy Kothinti, and Mounya Elhilali, “Temporal coding with magnitude-phase regularization for sound event detection,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{252337295, title = {Temporal coding with magnitude-phase regularization for sound event detection}, author = {{Sangwook Park} and {Sandeep Reddy Kothinti} and {Mounya Elhilali}}, year = 2022, month = {9}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/2c502d4f5eeb29bbf282d78d111cab0ed5d4cc00}, } - Jared Markowitz, Ryan W. Gardner, Ashley J. Llorens, R. Arora, and I-J. Wang, “A Risk-Sensitive Approach to Policy Optimization,” in AAAI Conference on Artificial Intelligence, 2022.
[BibTeX] [Link]@inproceedings{251710281, title = {A Risk-Sensitive Approach to Policy Optimization}, author = {{Jared Markowitz} and {Ryan W. Gardner} and {Ashley J. Llorens} and {R. Arora} and {I-J. Wang}}, year = 2022, month = {8}, booktitle = {AAAI Conference on Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/2a1b41221def527e17eb1ca04f4f32442fa09ba7}, } - Saurabh Kataria, J. Villalba, Laureano Moro-Vel’azquez, Piotr Żelasko, and N. Dehak, “Time-domain speech super-resolution with GAN based modeling for telephony speaker verification.” 2022.
[BibTeX] [Link]@inproceedings{252090174, title = {Time-domain speech super-resolution with GAN based modeling for telephony speaker verification}, author = {{Saurabh Kataria} and {J. Villalba} and {Laureano Moro-Vel'azquez} and {Piotr Żelasko} and {N. Dehak}}, year = 2022, month = {9}, booktitle = {}, url = {https://www.semanticscholar.org/paper/312a44c9d2d2719ca8d3eb22539edd215415229e}, } - Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao, Abu Awal Md Shoeb, Abubakar Abid, Adam Fisch, Adam R. Brown, Adam Santoro, Aditya Gupta, Adrià Garriga-Alonso, Agnieszka Kluska, Aitor Lewkowycz, Akshat Agarwal, Alethea Power, Alex Ray, Alex Warstadt, Alexander W. Kocurek, Ali Safaya, Ali Tazarv, Alice Xiang, Alicia Parrish, Allen Nie, Aman Hussain, Amanda Askell, Amanda Dsouza, Ambrose Slone, A. Rahane, Anantharaman S. Iyer, Anders Andreassen, Andrea Madotto, Andrea Santilli, Andreas Stuhlmuller, Andrew M. Dai, Andrew La, Andrew Kyle Lampinen, Andy Zou, Angela Jiang, Angelica Chen, Anh Vuong, Animesh Gupta, Anna Gottardi, Antonio Norelli, Anu Venkatesh, Arash Gholamidavoodi, Arfa Tabassum, Arul Menezes, Arun Kirubarajan, A. Mullokandov, Ashish Sabharwal, Austin Herrick, Avia Efrat, Aykut Erdem, Ayla Karakacs, B. R. Roberts, B. S. Loe, Barret Zoph, Bartlomiej Bojanowski, Batuhan Ozyurt, Behnam Hedayatnia, Behnam Neyshabur, Benjamin Inden, Benno Stein, Berk Ekmekci, Bill Yuchen Lin, B. Howald, Bryan Orinion, Cameron Diao, Cameron Dour, Catherine Stinson, Cedrick Argueta, C’esar Ferri Ram’irez, Chandan Singh, Charles Rathkopf, Chenlin Meng, Chitta Baral, Chiyu Wu, Chris Callison-Burch, Chris Waites, Christian Voigt, Christopher D. Manning, Christopher Potts, Cindy Ramirez, Clara E. Rivera, Clemencia Siro, Colin Raffel, Courtney Ashcraft, Cristina Garbacea, Damien Sileo, Daniel H Garrette, Dan Hendrycks, D. Kilman, Dan Roth, Daniel Freeman, Daniel Khashabi, Daniel Levy, D. Gonz’alez, Danielle R. Perszyk, Danny Hernandez, Danqi Chen, Daphne Ippolito, D. Gilboa, David Dohan, D. Drakard, David Jurgens, Debajyoti Datta, Deep Ganguli, Denis Emelin, D. Kleyko, Deniz Yuret, Derek Chen, Derek Tam, D. Hupkes, Diganta Misra, Dilyar Buzan, Dimitri Coelho Mollo, Diyi Yang, Dong-Ho Lee, Dylan Schrader, Ekaterina Shutova, E. D. Cubuk, Elad Segal, Eleanor Hagerman, Elizabeth Barnes, E. Donoway, Ellie Pavlick, E. Rodolà, Emma Lam, Eric Chu, Eric Tang, Erkut Erdem, Ernie Chang, Ethan A. Chi, Ethan Dyer, E. Jerzak, Ethan Kim, Eunice Engefu Manyasi, Evgenii Zheltonozhskii, Fanyue Xia, F. Siar, Fernando Mart’inez-Plumed, Francesca Happ’e, François Chollet, Frieda Rong, Gaurav Mishra, Genta Indra Winata, Gerard de Melo, Germán Kruszewski, Giambattista Parascandolo, Giorgio Mariani, Gloria Wang, Gonzalo Jaimovitch-L’opez, Gregor Betz, Guy Gur-Ari, Hana Galijasevic, Hannah Kim, Hannah Rashkin, Hannaneh Hajishirzi, Harsh Mehta, H. Bogar, Henry Shevlin, Hinrich Schutze, H. Yakura, Hongming Zhang, Hugh Mee Wong, Ian Ng, Isaac Noble, Jaap Jumelet, Jack Geissinger, John Kernion, Jacob Hilton, Jaehoon Lee, J. Fisac, James B. Simon, James Koppel, James Zheng, James Zou, Jan Koco’n, Jana Thompson, Janelle Wingfield, Jared Kaplan, Jarema Radom, Jascha Narain Sohl-Dickstein, Jason Phang, Jason Wei, J. Yosinski, Jekaterina Novikova, Jelle Bosscher, Jennifer Marsh, Jeremy Kim, Jeroen Taal, Jesse Engel, Jesujoba Oluwadara Alabi, Jiacheng Xu, Jiaming Song, Jillian Tang, Jane W Waweru, John Burden, John Miller, John U. Balis, Jonathan Batchelder, Jonathan Berant, Jorg Frohberg, Jos Rozen, J. Hernández-Orallo, Joseph Boudeman, Joseph Guerr, Joseph Jones, Joshua Tenenbaum, Joshua S. Rule, Joyce Chua, Kamil Kanclerz, Karen Livescu, K. Krauth, Karthik Gopalakrishnan, Katerina Ignatyeva, K. Markert, Kaustubh D. Dhole, Kevin Gimpel, Kevin Omondi, K. Mathewson, Kristen Chiafullo, Ksenia Shkaruta, K. Shridhar, Kyle McDonell, Kyle Richardson, Laria Reynolds, Leo Gao, Li Zhang, Liam Dugan, Lianhui Qin, Lidia Contreras-Ochando, Louis-Philippe Morency, Luca Moschella, Luca Lam, Lucy Noble, Ludwig Schmidt, Luheng He, Luis Oliveros Col’on, Luke Metz, Lutfi Kerem cSenel, Maarten Bosma, Maarten Sap, Maartje ter Hoeve, Maheen Farooqi, Manaal Faruqui, Mantas Mazeika, Marco Baturan, Marco Marelli, Marco Maru, Maria Jose Ram’irez Quintana, Marie Tolkiehn, Mario Giulianelli, Martha Lewis, Martin Potthast, Matthew L. Leavitt, Matthias Hagen, M. Schubert, Medina Baitemirova, Melody Arnaud, M. McElrath, Michael A. Yee, Michael Cohen, Michael Gu, Michael Ivanitskiy, Michael Starritt, M. Strube, Michal Swkedrowski, Michele Bevilacqua, Michihiro Yasunaga, Mihir Kale, Mike Cain, Mimee Xu, Mirac Suzgun, Mitch Walker, Monica Tiwari, Mohit Bansal, Moin Aminnaseri, Mor Geva, Mozhdeh Gheini, T. MukundVarma, Nanyun Peng, Nathan A. Chi, Nayeon Lee, Neta Gur-Ari Krakover, Nicholas Cameron, Nicholas Roberts, Nick Doiron, Nicole Martinez, Nikita Nangia, Niklas Deckers, Niklas Muennighoff, N. Keskar, Niveditha Iyer, Noah Constant, Noah Fiedel, Nuan Wen, Oliver Zhang, Omar Agha, Omar Elbaghdadi, Omer Levy, Owain Evans, Pablo Antonio Moreno Casares, P. Doshi, Pascale Fung, P. Liang, Paul Vicol, Pegah Alipoormolabashi, Peiyuan Liao, Percy Liang, Peter Chang, P. Eckersley, Phu Mon Htut, Pi-Bei Hwang, P. Milkowski, P. Patil, Pouya Pezeshkpour, P. Oli, Qiaozhu Mei, Qing Lyu, Qinlang Chen, Rabin Banjade, Rachel Etta Rudolph, Raefer Gabriel, Rahel Habacker, Ramon Risco, Raphael Milliere, Rhythm Garg, Richard Barnes, R. Saurous, Riku Arakawa, Robbe Raymaekers, Robert Frank, Rohan Sikand, Roman Novak, Roman Sitelew, Ronan Lebras, Rosanne Liu, Rowan Jacobs, Rui Zhang, R. Salakhutdinov, Ryan Chi, Ryan Lee, Ryan Stovall, Ryan Teehan, Rylan Yang, Sahib Singh, Saif M. Mohammad, Sajant Anand, Sam Dillavou, Sam Shleifer, Sam Wiseman, Samuel Gruetter, Samuel R. Bowman, S. Schoenholz, Sanghyun Han, Sanjeev Kwatra, Sarah A. Rous, Sarik Ghazarian, Sayan Ghosh, Sean Casey, Sebastian Bischoff, Sebastian Gehrmann, Sebastian Schuster, Sepideh Sadeghi, Shadi S. Hamdan, Sharon Zhou, Shashank Srivastava, Sherry Shi, Shikhar Singh, Shima Asaadi, Shixiang Shane Gu, Shubh Pachchigar, Shubham Toshniwal, Shyam Upadhyay, Shyamolima Debnath, Siamak Shakeri, Simon Thormeyer, S. Melzi, Siva Reddy, S. Makini, Soo-Hwan Lee, Spencer Torene, Sriharsha Hatwar, S. Dehaene, Stefan Divic, Stefano Ermon, Stella Biderman, Stephanie Lin, Stephen Prasad, Steven T Piantadosi, S. Shieber, Summer Misherghi, S. Kiritchenko, Swaroop Mishra, Tal Linzen, Tal Schuster, Tao Li, Tao Yu, Tariq Ali, Tatsunori Hashimoto, Te-Lin Wu, T. Desbordes, Theodore Rothschild, Thomas Phan, Tianle Wang, Tiberius Nkinyili, Timo Schick, T. Kornev, T. Tunduny, Tobias Gerstenberg, T. Chang, Trishala Neeraj, Tushar Khot, Tyler Shultz, Uri Shaham, Vedant Misra, Vera Demberg, Victoria Nyamai, Vikas Raunak, V. Ramasesh, Vinay Uday Prabhu, Vishakh Padmakumar, Vivek Srikumar, W. Fedus, W. Saunders, William Zhang, Wout Vossen, Xiang Ren, Xiaoyu Tong, Xinran Zhao, Xinyi Wu, Xudong Shen, Yadollah Yaghoobzadeh, Yair Lakretz, Yangqiu Song, Yasaman Bahri, Yejin Choi, Yichi Yang, Yiding Hao, Yifu Chen, Yonatan Belinkov, Yu Hou, Yu Hou, Yuntao Bai, Zachary Seid, Zhuoye Zhao, Z. Wang, Zijie J. Wang, Zirui Wang, and Ziyi Wu, “Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models.” 2022.
[BibTeX] [Link]@inproceedings{263625818, title = {Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models}, author = {{Aarohi Srivastava} and {Abhinav Rastogi} and {Abhishek Rao} and {Abu Awal Md Shoeb} and {Abubakar Abid} and {Adam Fisch} and {Adam R. Brown} and {Adam Santoro} and {Aditya Gupta} and {Adrià Garriga-Alonso} and {Agnieszka Kluska} and {Aitor Lewkowycz} and {Akshat Agarwal} and {Alethea Power} and {Alex Ray} and {Alex Warstadt} and {Alexander W. Kocurek} and {Ali Safaya} and {Ali Tazarv} and {Alice Xiang} and {Alicia Parrish} and {Allen Nie} and {Aman Hussain} and {Amanda Askell} and {Amanda Dsouza} and {Ambrose Slone} and {A. Rahane} and {Anantharaman S. Iyer} and {Anders Andreassen} and {Andrea Madotto} and {Andrea Santilli} and {Andreas Stuhlmuller} and {Andrew M. Dai} and {Andrew La} and {Andrew Kyle Lampinen} and {Andy Zou} and {Angela Jiang} and {Angelica Chen} and {Anh Vuong} and {Animesh Gupta} and {Anna Gottardi} and {Antonio Norelli} and {Anu Venkatesh} and {Arash Gholamidavoodi} and {Arfa Tabassum} and {Arul Menezes} and {Arun Kirubarajan} and {A. Mullokandov} and {Ashish Sabharwal} and {Austin Herrick} and {Avia Efrat} and {Aykut Erdem} and {Ayla Karakacs} and {B. R. Roberts} and {B. S. Loe} and {Barret Zoph} and {Bartlomiej Bojanowski} and {Batuhan Ozyurt} and {Behnam Hedayatnia} and {Behnam Neyshabur} and {Benjamin Inden} and {Benno Stein} and {Berk Ekmekci} and {Bill Yuchen Lin} and {B. Howald} and {Bryan Orinion} and {Cameron Diao} and {Cameron Dour} and {Catherine Stinson} and {Cedrick Argueta} and {C'esar Ferri Ram'irez} and {Chandan Singh} and {Charles Rathkopf} and {Chenlin Meng} and {Chitta Baral} and {Chiyu Wu} and {Chris Callison-Burch} and {Chris Waites} and {Christian Voigt} and {Christopher D. Manning} and {Christopher Potts} and {Cindy Ramirez} and {Clara E. Rivera} and {Clemencia Siro} and {Colin Raffel} and {Courtney Ashcraft} and {Cristina Garbacea} and {Damien Sileo} and {Daniel H Garrette} and {Dan Hendrycks} and {D. Kilman} and {Dan Roth} and {Daniel Freeman} and {Daniel Khashabi} and {Daniel Levy} and {D. Gonz'alez} and {Danielle R. Perszyk} and {Danny Hernandez} and {Danqi Chen} and {Daphne Ippolito} and {D. Gilboa} and {David Dohan} and {D. Drakard} and {David Jurgens} and {Debajyoti Datta} and {Deep Ganguli} and {Denis Emelin} and {D. Kleyko} and {Deniz Yuret} and {Derek Chen} and {Derek Tam} and {D. Hupkes} and {Diganta Misra} and {Dilyar Buzan} and {Dimitri Coelho Mollo} and {Diyi Yang} and {Dong-Ho Lee} and {Dylan Schrader} and {Ekaterina Shutova} and {E. D. Cubuk} and {Elad Segal} and {Eleanor Hagerman} and {Elizabeth Barnes} and {E. Donoway} and {Ellie Pavlick} and {E. Rodolà} and {Emma Lam} and {Eric Chu} and {Eric Tang} and {Erkut Erdem} and {Ernie Chang} and {Ethan A. Chi} and {Ethan Dyer} and {E. Jerzak} and {Ethan Kim} and {Eunice Engefu Manyasi} and {Evgenii Zheltonozhskii} and {Fanyue Xia} and {F. Siar} and {Fernando Mart'inez-Plumed} and {Francesca Happ'e} and {François Chollet} and {Frieda Rong} and {Gaurav Mishra} and {Genta Indra Winata} and {Gerard de Melo} and {Germán Kruszewski} and {Giambattista Parascandolo} and {Giorgio Mariani} and {Gloria Wang} and {Gonzalo Jaimovitch-L'opez} and {Gregor Betz} and {Guy Gur-Ari} and {Hana Galijasevic} and {Hannah Kim} and {Hannah Rashkin} and {Hannaneh Hajishirzi} and {Harsh Mehta} and {H. Bogar} and {Henry Shevlin} and {Hinrich Schutze} and {H. Yakura} and {Hongming Zhang} and {Hugh Mee Wong} and {Ian Ng} and {Isaac Noble} and {Jaap Jumelet} and {Jack Geissinger} and {John Kernion} and {Jacob Hilton} and {Jaehoon Lee} and {J. Fisac} and {James B. Simon} and {James Koppel} and {James Zheng} and {James Zou} and {Jan Koco'n} and {Jana Thompson} and {Janelle Wingfield} and {Jared Kaplan} and {Jarema Radom} and {Jascha Narain Sohl-Dickstein} and {Jason Phang} and {Jason Wei} and {J. Yosinski} and {Jekaterina Novikova} and {Jelle Bosscher} and {Jennifer Marsh} and {Jeremy Kim} and {Jeroen Taal} and {Jesse Engel} and {Jesujoba Oluwadara Alabi} and {Jiacheng Xu} and {Jiaming Song} and {Jillian Tang} and {Jane W Waweru} and {John Burden} and {John Miller} and {John U. Balis} and {Jonathan Batchelder} and {Jonathan Berant} and {Jorg Frohberg} and {Jos Rozen} and {J. Hernández-Orallo} and {Joseph Boudeman} and {Joseph Guerr} and {Joseph Jones} and {Joshua Tenenbaum} and {Joshua S. Rule} and {Joyce Chua} and {Kamil Kanclerz} and {Karen Livescu} and {K. Krauth} and {Karthik Gopalakrishnan} and {Katerina Ignatyeva} and {K. Markert} and {Kaustubh D. Dhole} and {Kevin Gimpel} and {Kevin Omondi} and {K. Mathewson} and {Kristen Chiafullo} and {Ksenia Shkaruta} and {K. Shridhar} and {Kyle McDonell} and {Kyle Richardson} and {Laria Reynolds} and {Leo Gao} and {Li Zhang} and {Liam Dugan} and {Lianhui Qin} and {Lidia Contreras-Ochando} and {Louis-Philippe Morency} and {Luca Moschella} and {Luca Lam} and {Lucy Noble} and {Ludwig Schmidt} and {Luheng He} and {Luis Oliveros Col'on} and {Luke Metz} and {Lutfi Kerem cSenel} and {Maarten Bosma} and {Maarten Sap} and {Maartje ter Hoeve} and {Maheen Farooqi} and {Manaal Faruqui} and {Mantas Mazeika} and {Marco Baturan} and {Marco Marelli} and {Marco Maru} and {Maria Jose Ram'irez Quintana} and {Marie Tolkiehn} and {Mario Giulianelli} and {Martha Lewis} and {Martin Potthast} and {Matthew L. Leavitt} and {Matthias Hagen} and {M. Schubert} and {Medina Baitemirova} and {Melody Arnaud} and {M. McElrath} and {Michael A. Yee} and {Michael Cohen} and {Michael Gu} and {Michael Ivanitskiy} and {Michael Starritt} and {M. Strube} and {Michal Swkedrowski} and {Michele Bevilacqua} and {Michihiro Yasunaga} and {Mihir Kale} and {Mike Cain} and {Mimee Xu} and {Mirac Suzgun} and {Mitch Walker} and {Monica Tiwari} and {Mohit Bansal} and {Moin Aminnaseri} and {Mor Geva} and {Mozhdeh Gheini} and {T. MukundVarma} and {Nanyun Peng} and {Nathan A. Chi} and {Nayeon Lee} and {Neta Gur-Ari Krakover} and {Nicholas Cameron} and {Nicholas Roberts} and {Nick Doiron} and {Nicole Martinez} and {Nikita Nangia} and {Niklas Deckers} and {Niklas Muennighoff} and {N. Keskar} and {Niveditha Iyer} and {Noah Constant} and {Noah Fiedel} and {Nuan Wen} and {Oliver Zhang} and {Omar Agha} and {Omar Elbaghdadi} and {Omer Levy} and {Owain Evans} and {Pablo Antonio Moreno Casares} and {P. Doshi} and {Pascale Fung} and {P. Liang} and {Paul Vicol} and {Pegah Alipoormolabashi} and {Peiyuan Liao} and {Percy Liang} and {Peter Chang} and {P. Eckersley} and {Phu Mon Htut} and {Pi-Bei Hwang} and {P. Milkowski} and {P. Patil} and {Pouya Pezeshkpour} and {P. Oli} and {Qiaozhu Mei} and {Qing Lyu} and {Qinlang Chen} and {Rabin Banjade} and {Rachel Etta Rudolph} and {Raefer Gabriel} and {Rahel Habacker} and {Ramon Risco} and {Raphael Milliere} and {Rhythm Garg} and {Richard Barnes} and {R. Saurous} and {Riku Arakawa} and {Robbe Raymaekers} and {Robert Frank} and {Rohan Sikand} and {Roman Novak} and {Roman Sitelew} and {Ronan Lebras} and {Rosanne Liu} and {Rowan Jacobs} and {Rui Zhang} and {R. Salakhutdinov} and {Ryan Chi} and {Ryan Lee} and {Ryan Stovall} and {Ryan Teehan} and {Rylan Yang} and {Sahib Singh} and {Saif M. Mohammad} and {Sajant Anand} and {Sam Dillavou} and {Sam Shleifer} and {Sam Wiseman} and {Samuel Gruetter} and {Samuel R. Bowman} and {S. Schoenholz} and {Sanghyun Han} and {Sanjeev Kwatra} and {Sarah A. Rous} and {Sarik Ghazarian} and {Sayan Ghosh} and {Sean Casey} and {Sebastian Bischoff} and {Sebastian Gehrmann} and {Sebastian Schuster} and {Sepideh Sadeghi} and {Shadi S. Hamdan} and {Sharon Zhou} and {Shashank Srivastava} and {Sherry Shi} and {Shikhar Singh} and {Shima Asaadi} and {Shixiang Shane Gu} and {Shubh Pachchigar} and {Shubham Toshniwal} and {Shyam Upadhyay} and {Shyamolima Debnath} and {Siamak Shakeri} and {Simon Thormeyer} and {S. Melzi} and {Siva Reddy} and {S. Makini} and {Soo-Hwan Lee} and {Spencer Torene} and {Sriharsha Hatwar} and {S. Dehaene} and {Stefan Divic} and {Stefano Ermon} and {Stella Biderman} and {Stephanie Lin} and {Stephen Prasad} and {Steven T Piantadosi} and {S. Shieber} and {Summer Misherghi} and {S. Kiritchenko} and {Swaroop Mishra} and {Tal Linzen} and {Tal Schuster} and {Tao Li} and {Tao Yu} and {Tariq Ali} and {Tatsunori Hashimoto} and {Te-Lin Wu} and {T. Desbordes} and {Theodore Rothschild} and {Thomas Phan} and {Tianle Wang} and {Tiberius Nkinyili} and {Timo Schick} and {T. Kornev} and {T. Tunduny} and {Tobias Gerstenberg} and {T. Chang} and {Trishala Neeraj} and {Tushar Khot} and {Tyler Shultz} and {Uri Shaham} and {Vedant Misra} and {Vera Demberg} and {Victoria Nyamai} and {Vikas Raunak} and {V. Ramasesh} and {Vinay Uday Prabhu} and {Vishakh Padmakumar} and {Vivek Srikumar} and {W. Fedus} and {W. Saunders} and {William Zhang} and {Wout Vossen} and {Xiang Ren} and {Xiaoyu Tong} and {Xinran Zhao} and {Xinyi Wu} and {Xudong Shen} and {Yadollah Yaghoobzadeh} and {Yair Lakretz} and {Yangqiu Song} and {Yasaman Bahri} and {Yejin Choi} and {Yichi Yang} and {Yiding Hao} and {Yifu Chen} and {Yonatan Belinkov} and {Yu Hou} and {Yu Hou} and {Yuntao Bai} and {Zachary Seid} and {Zhuoye Zhao} and {Z. Wang} and {Zijie J. Wang} and {Zirui Wang} and {Ziyi Wu}}, year = 2022, month = {6}, booktitle = {}, url = {https://www.semanticscholar.org/paper/bd1331b233e84bab7eba503abc60b31ac08e7881}, } - Suzanna Sia, Kokil Jaidka, Niyati Chayya, and Kevin Duh, “Modeling Constraints Can Identify Winning Arguments in Multi-Party Interactions (Student Abstract),” in AAAI Conference on Artificial Intelligence, 2022.
[BibTeX] [Link]@inproceedings{250298720, title = {Modeling Constraints Can Identify Winning Arguments in Multi-Party Interactions (Student Abstract)}, author = {{Suzanna Sia} and {Kokil Jaidka} and {Niyati Chayya} and {Kevin Duh}}, year = 2022, month = {6}, booktitle = {AAAI Conference on Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/da88a7e2b2187fc230b61f36752dbf396be9ce32}, } - Sucheng Ren, Huiyu Wang, Zhengqi Gao, Shengfeng He, A. Yuille, Yuyin Zhou, and Cihang Xie, “A Simple Data Mixing Prior for Improving Self-Supervised Learning,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{249037240, title = {A Simple Data Mixing Prior for Improving Self-Supervised Learning}, author = {{Sucheng Ren} and {Huiyu Wang} and {Zhengqi Gao} and {Shengfeng He} and {A. Yuille} and {Yuyin Zhou} and {Cihang Xie}}, year = 2022, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/5ed5dcb0763af9e6283dcdcf4af75248d9d19c95}, } - Hexin Liu, Leibny Paola Garcia Perera, Andy W. H. Khong, S. Styles, and S. Khudanpur, “PHO-LID: A Unified Model Incorporating Acoustic-Phonetic and Phonotactic Information for Language Identification,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{247618802, title = {PHO-LID: A Unified Model Incorporating Acoustic-Phonetic and Phonotactic Information for Language Identification}, author = {{Hexin Liu} and {Leibny Paola Garcia Perera} and {Andy W. H. Khong} and {S. Styles} and {S. Khudanpur}}, year = 2022, month = {3}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/3f7542a6f77db123632ac723ab49f5a62f6184e3}, } - Shao-Yuan Lo, Wei Wang, Jim Thomas, Jingjing Zheng, Vishal M. Patel, and Cheng-Hao Kuo, “Learning Feature Decomposition for Domain Adaptive Monocular Depth Estimation,” in IEEE/RJS International Conference on Intelligent RObots and Systems, 2022.
[BibTeX] [Link]@inproceedings{251224367, title = {Learning Feature Decomposition for Domain Adaptive Monocular Depth Estimation}, author = {{Shao-Yuan Lo} and {Wei Wang} and {Jim Thomas} and {Jingjing Zheng} and {Vishal M. Patel} and {Cheng-Hao Kuo}}, year = 2022, month = {7}, booktitle = {IEEE/RJS International Conference on Intelligent RObots and Systems}, url = {https://www.semanticscholar.org/paper/c28582e042a0bb482517ef844d5a3a6794f994f6}, } - Magdalena Rybicka, J. Villalba, N. Dehak, and K. Kowalczyk, “End-to-End Neural Speaker Diarization with an Iterative Refinement of Non-Autoregressive Attention-based Attractors,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{252346611, title = {End-to-End Neural Speaker Diarization with an Iterative Refinement of Non-Autoregressive Attention-based Attractors}, author = {{Magdalena Rybicka} and {J. Villalba} and {N. Dehak} and {K. Kowalczyk}}, year = 2022, month = {9}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/916cfa98c48af9931559fe0d8953bcaf7bdf7f2c}, } - Yutong Bai, Xinlei Chen, Alexander Kirillov, A. Yuille, and A. Berg, “Point-Level Region Contrast for Object Detection Pre-Training,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{246680149, title = {Point-Level Region Contrast for Object Detection Pre-Training}, author = {{Yutong Bai} and {Xinlei Chen} and {Alexander Kirillov} and {A. Yuille} and {A. Berg}}, year = 2022, month = {2}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/7d692139562f9679a3694e1d408b00bd8259b6f1}, } - Drew Prinster, Anqi Liu, and S. Saria, “JAWS: Predictive Inference Under Covariate Shift,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{251018520, title = {JAWS: Predictive Inference Under Covariate Shift}, author = {{Drew Prinster} and {Anqi Liu} and {S. Saria}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/e9b0db3dae9050413e3eda2861acf82bff41624b}, } - Chan Young Park, Julia Mendelsohn, Anjalie Field, and Yulia Tsvetkov, “VoynaSlov: A Data Set of Russian Social Media Activity during the 2022 Ukraine-Russia War,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249063004, title = {VoynaSlov: A Data Set of Russian Social Media Activity during the 2022 Ukraine-Russia War}, author = {{Chan Young Park} and {Julia Mendelsohn} and {Anjalie Field} and {Yulia Tsvetkov}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/215a6f2b4c206975f59d81c0c9f45cfe935a85e9}, } - Matthew Maciejewski, Jing Shi, Shinji Watanabe, and S. Khudanpur, “A dilemma of ground truth in noisy speech separation and an approach to lessen the impact of imperfect training data,” in Computer Speech and Language, 2022.
[BibTeX] [Link]@inproceedings{249386629, title = {A dilemma of ground truth in noisy speech separation and an approach to lessen the impact of imperfect training data}, author = {{Matthew Maciejewski} and {Jing Shi} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2022, month = {6}, booktitle = {Computer Speech and Language}, url = {https://www.semanticscholar.org/paper/e3b6ab2d2e1a0e734bf505fbb34dc6fe723ab37e}, } - K. Henry, R. Kornfield, A. Sridharan, Robert C. Linton, Catherine Groh, Tony Wang, Albert W. Wu, Bilge Mutlu, and S. Saria, “Human–machine teaming is key to AI adoption: clinicians’ experiences with a deployed machine learning system,” in npj Digital Medicine, 2022.
[BibTeX] [Link]@inproceedings{250703036, title = {Human–machine teaming is key to AI adoption: clinicians’ experiences with a deployed machine learning system}, author = {{K. Henry} and {R. Kornfield} and {A. Sridharan} and {Robert C. Linton} and {Catherine Groh} and {Tony Wang} and {Albert W. Wu} and {Bilge Mutlu} and {S. Saria}}, year = 2022, month = {7}, booktitle = {npj Digital Medicine}, url = {https://www.semanticscholar.org/paper/22ea0aa6c750d327529053d66e4f0a9457485402}, } - Saksham Suri, Saketh Rambhatla, R. Chellappa, and Abhinav Shrivastava, “R-SSL: Region based Semi-Supervised Learning for Sparsely Annotated Object Detection.” 2022.
[BibTeX] [Link]@inproceedings{255750913, title = {R-SSL: Region based Semi-Supervised Learning for Sparsely Annotated Object Detection}, author = {{Saksham Suri} and {Saketh Rambhatla} and {R. Chellappa} and {Abhinav Shrivastava}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/e2e159205030b9d3e3d742b4bdbebd7e94201d3f}, } - Pirazh Khorramshahi, Vineet Shenoy, M. Pack, and R. Chellappa, “Scalable and Real-time Multi-Camera Vehicle Detection, Re-Identification, and Tracking,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248218560, title = {Scalable and Real-time Multi-Camera Vehicle Detection, Re-Identification, and Tracking}, author = {{Pirazh Khorramshahi} and {Vineet Shenoy} and {M. Pack} and {R. Chellappa}}, year = 2022, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/0babd241088a1d84dec824c9749c93a3e20fd583}, } - R. Wicks and K. Duh, “The Effects of Language Token Prefixing for Multilingual Machine Translation,” in Proceedings of the 2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), Online only, 2022, p. 148–153.
[BibTeX] [Abstract] [Link]
Machine translation traditionally refers to translating from a single source language into a single target language. In recent years, the field has moved towards large neural models either translating from or into many languages. The model must be correctly cued to translate into the correct target language. This is typically done by prefixing language tokens onto the source or target sequence. The location and content of the prefix can vary and many use different approaches without much justification towards one approach or another. As a guidance to future researchers and directions for future work, we present a series of experiments that show how the positioning and type of a target language prefix token effects translation performance. We show that source side prefixes improve performance. Further, we find that the best language information to denote via tokens depends on the supported language set.
@inproceedings{wicks-duh-2022-effects, title = "The Effects of Language Token Prefixing for Multilingual Machine Translation", author = "Wicks, Rachel and Duh, Kevin", editor = "He, Yulan and Ji, Heng and Li, Sujian and Liu, Yang and Chang, Chua-Hui", booktitle = "Proceedings of the 2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)", month = nov, year = "2022", address = "Online only", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.aacl-short.19", pages = "148--153", abstract = "Machine translation traditionally refers to translating from a single source language into a single target language. In recent years, the field has moved towards large neural models either translating from or into many languages. The model must be correctly cued to translate into the correct target language. This is typically done by prefixing language tokens onto the source or target sequence. The location and content of the prefix can vary and many use different approaches without much justification towards one approach or another. As a guidance to future researchers and directions for future work, we present a series of experiments that show how the positioning and type of a target language prefix token effects translation performance. We show that source side prefixes improve performance. Further, we find that the best language information to denote via tokens depends on the supported language set.", } - Cheng Peng and R. Chellappa, “PDRF: Progressively Deblurring Radiance Field for Fast and Robust Scene Reconstruction from Blurry Images,” in AAAI Conference on Artificial Intelligence, 2022.
[BibTeX] [Link]@inproceedings{251622408, title = {PDRF: Progressively Deblurring Radiance Field for Fast and Robust Scene Reconstruction from Blurry Images}, author = {{Cheng Peng} and {R. Chellappa}}, year = 2022, month = {8}, booktitle = {AAAI Conference on Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/c900f690fdab5d17b0253d4362e7f1a7d9d2d495}, } - Pengfei Guo, Dong Yang, Ali Hatamizadeh, An Xu, Ziyue Xu, Wenqi Li, Can Zhao, Daguang Xu, S. Harmon, E. Turkbey, B. Turkbey, B. Wood, F. Patella, Elvira Stellato, G. Carrafiello, Vishal M. Patel, and H. Roth, “Auto-FedRL: Federated Hyperparameter Optimization for Multi-institutional Medical Image Segmentation,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{247447734, title = {Auto-FedRL: Federated Hyperparameter Optimization for Multi-institutional Medical Image Segmentation}, author = {{Pengfei Guo} and {Dong Yang} and {Ali Hatamizadeh} and {An Xu} and {Ziyue Xu} and {Wenqi Li} and {Can Zhao} and {Daguang Xu} and {S. Harmon} and {E. Turkbey} and {B. Turkbey} and {B. Wood} and {F. Patella} and {Elvira Stellato} and {G. Carrafiello} and {Vishal M. Patel} and {H. Roth}}, year = 2022, month = {3}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/ea8889c3bbca75fcdd71ba60068df014dfb7d861}, } - Subhro Roy, Sam Thomson, Tongfei Chen, Richard Shin, Adam Pauls, J. Eisner, and Benjamin Van Durme, “BenchCLAMP: A Benchmark for Evaluating Language Models on Semantic Parsing,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249926634, title = {BenchCLAMP: A Benchmark for Evaluating Language Models on Semantic Parsing}, author = {{Subhro Roy} and {Sam Thomson} and {Tongfei Chen} and {Richard Shin} and {Adam Pauls} and {J. Eisner} and {Benjamin Van Durme}}, year = 2022, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/132751a80632e80a90d7c3d3cd8a361f48fdb9b4}, } - Amir Alipour-Fanid, Monireh Dabaghchian, R. Arora, and K. Zeng, “Multiuser Scheduling in Centralized Cognitive Radio Networks: A Multi-Armed Bandit Approach,” in IEEE Transactions on Cognitive Communications and Networking, 2022.
[BibTeX] [Link]@inproceedings{246595318, title = {Multiuser Scheduling in Centralized Cognitive Radio Networks: A Multi-Armed Bandit Approach}, author = {{Amir Alipour-Fanid} and {Monireh Dabaghchian} and {R. Arora} and {K. Zeng}}, year = 2022, month = {6}, booktitle = {IEEE Transactions on Cognitive Communications and Networking}, url = {https://www.semanticscholar.org/paper/ad0c8cc0a80c5873591e62ca9f47fa21b631c35f}, } - Mo Zhou and Vishal M. Patel, “On Trace of PGD-Like Adversarial Attacks,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248887310, title = {On Trace of PGD-Like Adversarial Attacks}, author = {{Mo Zhou} and {Vishal M. Patel}}, year = 2022, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/90d02089aaf88b621880a036a2cc4c5924f7102c}, } - Daniel E Park, Nora L. Watson, Christopher Focht, D. Feikin, Laura L Hammit, W. A. Brooks, S. Howie, K. Kotloff, O. Levine, S. Madhi, D. Murdoch, K. O’Brien, J. Scott, D. Thea, Tussanee Amorninthapichet, Juliet O. Awori, C. Bunthi, B. Ebruke, Mounya Elhilali, Melissa M Higdon, L. Hossain, Y. Jahan, D. Moore, J. Mulindwa, L. Mwananyanda, Sathapana Naorat, Christine Prosperi, S. Thamthitiwat, C. Verwey, K. Jablonski, M. Power, H. Young, M. Deloria Knoll, and E. McCollum, “Digitally recorded and remotely classified lung auscultation compared with conventional stethoscope classifications among children aged 1–59 months enrolled in the Pneumonia Etiology Research for Child Health (PERCH) case–control study,” in BMJ Open Respiratory Research, 2022.
[BibTeX] [Link]@inproceedings{248832550, title = {Digitally recorded and remotely classified lung auscultation compared with conventional stethoscope classifications among children aged 1–59 months enrolled in the Pneumonia Etiology Research for Child Health (PERCH) case–control study}, author = {{Daniel E Park} and {Nora L. Watson} and {Christopher Focht} and {D. Feikin} and {Laura L Hammit} and {W. A. Brooks} and {S. Howie} and {K. Kotloff} and {O. Levine} and {S. Madhi} and {D. Murdoch} and {K. O'Brien} and {J. Scott} and {D. Thea} and {Tussanee Amorninthapichet} and {Juliet O. Awori} and {C. Bunthi} and {B. Ebruke} and {Mounya Elhilali} and {Melissa M Higdon} and {L. Hossain} and {Y. Jahan} and {D. Moore} and {J. Mulindwa} and {L. Mwananyanda} and {Sathapana Naorat} and {Christine Prosperi} and {S. Thamthitiwat} and {C. Verwey} and {K. Jablonski} and {M. Power} and {H. Young} and {M. Deloria Knoll} and {E. McCollum}}, year = 2022, month = {5}, booktitle = {BMJ Open Respiratory Research}, url = {https://www.semanticscholar.org/paper/dfedb313d8718de8aa162813060af3e24e8cbe28}, } - Shraman Pramanick, E. Nowara, Joshua Gleason, C. Castillo, and R. Chellappa, “Where in the World is this Image? Transformer-based Geo-localization in the Wild,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{248476325, title = {Where in the World is this Image? Transformer-based Geo-localization in the Wild}, author = {{Shraman Pramanick} and {E. Nowara} and {Joshua Gleason} and {C. Castillo} and {R. Chellappa}}, year = 2022, month = {4}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/1889dfb7c30f2b9f8e9d4026ca71ec10caa449af}, } - Xiang Xiang, Yuwen Tan, Qian Wan, Jing Ma, A. Yuille, and Gregory D. Hager, “Coarse-To-Fine Incremental Few-Shot Learning – Appendix.” 2022.
[BibTeX] [Link]@inproceedings{253540960, title = {Coarse-To-Fine Incremental Few-Shot Learning - Appendix}, author = {{Xiang Xiang} and {Yuwen Tan} and {Qian Wan} and {Jing Ma} and {A. Yuille} and {Gregory D. Hager}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/4656e23147a7bf6cce8ef8702324da910d004bb4}, } - Hexin Liu, Leibny Paola Garcia Perera, Andy W. H. Khong, J. Dauwels, S. Styles, and S. Khudanpur, “Enhance Language Identification using Dual-mode Model with Knowledge Distillation,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{247291930, title = {Enhance Language Identification using Dual-mode Model with Knowledge Distillation}, author = {{Hexin Liu} and {Leibny Paola Garcia Perera} and {Andy W. H. Khong} and {J. Dauwels} and {S. Styles} and {S. Khudanpur}}, year = 2022, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b238287cad4bf831f1f7600207e967b95620017d}, } - Zili Huang, Shinji Watanabe, Shu-wen Yang, Leibny Paola García-Perera, and S. Khudanpur, “Investigating Self-Supervised Learning for Speech Enhancement and Separation,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{247450789, title = {Investigating Self-Supervised Learning for Speech Enhancement and Separation}, author = {{Zili Huang} and {Shinji Watanabe} and {Shu-wen Yang} and {Leibny Paola García-Perera} and {S. Khudanpur}}, year = 2022, month = {3}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/d5634a21b3727258822b78f5c5ababf7261a5c79}, } - W. G. C. Bandara and Vishal M. Patel, “HyperTransformer: A Textural and Spectral Feature Fusion Transformer for Pansharpening,” in Computer Vision and Pattern Recognition, 2022.
[BibTeX] [Link]@inproceedings{247244749, title = {HyperTransformer: A Textural and Spectral Feature Fusion Transformer for Pansharpening}, author = {{W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/0477780c61e668c47630ae1cd74cee55c2493b5f}, } - Pirazh Khorramshahi, Vineet Shenoy, and R. Chellappa, “Scalable Vehicle Re-Identification via Self-Supervision,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248811372, title = {Scalable Vehicle Re-Identification via Self-Supervision}, author = {{Pirazh Khorramshahi} and {Vineet Shenoy} and {R. Chellappa}}, year = 2022, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/9d69f0b6c916ac36e2bf28491d27c653eae245cd}, } - B. Vasey, M. Nagendran, Bruce Campbell, D. Clifton, Gary S. Collins, Spiros C. Denaxas, A. Denniston, L. Faes, B. Geerts, Mudathir Ibrahim, Xiaoxuan Liu, B. Mateen, P. Mathur, M. Mccradden, L. Morgan, Johan Ordish, Campbell Rogers, S. Saria, D. Ting, P. Watkinson, W. Weber, P. Wheatstone, and P. McCulloch, “Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI,” in Nature Network Boston, 2022.
[BibTeX] [Link]@inproceedings{248890002, title = {Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI}, author = {{B. Vasey} and {M. Nagendran} and {Bruce Campbell} and {D. Clifton} and {Gary S. Collins} and {Spiros C. Denaxas} and {A. Denniston} and {L. Faes} and {B. Geerts} and {Mudathir Ibrahim} and {Xiaoxuan Liu} and {B. Mateen} and {P. Mathur} and {M. Mccradden} and {L. Morgan} and {Johan Ordish} and {Campbell Rogers} and {S. Saria} and {D. Ting} and {P. Watkinson} and {W. Weber} and {P. Wheatstone} and {P. McCulloch}}, year = 2022, month = {5}, booktitle = {Nature Network Boston}, url = {https://www.semanticscholar.org/paper/83b6a76ba5112d27bdbfca3efd2ed918d8e73db5}, } - Vikas Raunak, Matt Post, and Arul Menezes, “SALTED: A Framework for SAlient Long-Tail Translation Error Detection,” in Conference on Empirical Methods in Natural Language Processing, 2022.
[BibTeX] [Link]@inproceedings{248965209, title = {SALTED: A Framework for SAlient Long-Tail Translation Error Detection}, author = {{Vikas Raunak} and {Matt Post} and {Arul Menezes}}, year = 2022, month = {5}, booktitle = {Conference on Empirical Methods in Natural Language Processing}, url = {https://www.semanticscholar.org/paper/a349bcb86ba80ef543e5deaadbb7e0ff5daef5e7}, } - R. Yasarla, Renliang Weng, Wongun Choi, Vishal M. Patel, and Amir Sadeghian, “3SD: Self-Supervised Saliency Detection With No Labels,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{247318765, title = {3SD: Self-Supervised Saliency Detection With No Labels}, author = {{R. Yasarla} and {Renliang Weng} and {Wongun Choi} and {Vishal M. Patel} and {Amir Sadeghian}}, year = 2022, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/2a78e1c0412cbcc851ba60224c15c501debe2049}, } - Bruce Y Lee, J. Ordovás, E. Parks, Cheryl Anderson, A. Barabási, S. Clinton, K. de la Haye, V. Duffy, P. Franks, Elizabeth M Ginexi, K. Hammond, Erin C. Hanlon, Michael Hittle, E. Ho, A. Horn, R. Isaacson, P. Mabry, S. Malone, Corby K. Martin, J. Mattei, S. Meydani, Lorene M. Nelson, M. Neuhouser, Brendan Parent, N. Pronk, H. Roche, S. Saria, F. Scheer, E. Segal, M. Sevick, T. Spector, Linda Van Horn, K. Varady, V. S. Voruganti, and Marie F Martinez, “Research gaps and opportunities in precision nutrition: an NIH workshop report.,” in American Journal of Clinical Nutrition, 2022.
[BibTeX] [Link]@inproceedings{252045737, title = {Research gaps and opportunities in precision nutrition: an NIH workshop report.}, author = {{Bruce Y Lee} and {J. Ordovás} and {E. Parks} and {Cheryl Anderson} and {A. Barabási} and {S. Clinton} and {K. de la Haye} and {V. Duffy} and {P. Franks} and {Elizabeth M Ginexi} and {K. Hammond} and {Erin C. Hanlon} and {Michael Hittle} and {E. Ho} and {A. Horn} and {R. Isaacson} and {P. Mabry} and {S. Malone} and {Corby K. Martin} and {J. Mattei} and {S. Meydani} and {Lorene M. Nelson} and {M. Neuhouser} and {Brendan Parent} and {N. Pronk} and {H. Roche} and {S. Saria} and {F. Scheer} and {E. Segal} and {M. Sevick} and {T. Spector} and {Linda Van Horn} and {K. Varady} and {V. S. Voruganti} and {Marie F Martinez}}, year = 2022, month = {9}, booktitle = {American Journal of Clinical Nutrition}, url = {https://www.semanticscholar.org/paper/31b65b7ccc0ed5ba975753c8b0ba8da8df28a09c}, } - Daniel Khashabi, Yeganeh Kordi, and Hannaneh Hajishirzi, “UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Training,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{247154787, title = {UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Training}, author = {{Daniel Khashabi} and {Yeganeh Kordi} and {Hannaneh Hajishirzi}}, year = 2022, month = {2}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/5b44101b2372a33ec06e15ce4d20ad9a15518325}, } - Feng Wang, Huiyu Wang, Chen Wei, A. Yuille, and Wei Shen, “CP2: Copy-Paste Contrastive Pretraining for Semantic Segmentation,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{247596632, title = {CP2: Copy-Paste Contrastive Pretraining for Semantic Segmentation}, author = {{Feng Wang} and {Huiyu Wang} and {Chen Wei} and {A. Yuille} and {Wei Shen}}, year = 2022, month = {3}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/3eb748f6279de5cfc582b3179bd1012bbd95614e}, } - Kelly Marchisio, Conghao Xiong, and Philipp Koehn, “Embedding-Enhanced GIZA++: Improving Low-Resource Word Alignment Using Embeddings,” in Conference of the Association for Machine Translation in the Americas, 2022.
[BibTeX] [Link]@inproceedings{252186400, title = {Embedding-Enhanced GIZA++: Improving Low-Resource Word Alignment Using Embeddings}, author = {{Kelly Marchisio} and {Conghao Xiong} and {Philipp Koehn}}, year = 2022, booktitle = {Conference of the Association for Machine Translation in the Americas}, url = {https://www.semanticscholar.org/paper/4768c7f83f1c4fbb4fd98d9b4237ab483a8bc4b2}, } - Martin Sustek, Samik Sadhu, and H. Hermansky, “Dealing with Unknowns in Continual Learning for End-to-end Automatic Speech Recognition,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{252337895, title = {Dealing with Unknowns in Continual Learning for End-to-end Automatic Speech Recognition}, author = {{Martin Sustek} and {Samik Sadhu} and {H. Hermansky}}, year = 2022, month = {9}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/dea2103e2b666413670b3f5c81a2e3ca318ea2d4}, } - Yiqun Mei, Pengfei Guo, and Vishal M. Patel, “Supplementary File: Escaping Data Scarcity for High-Resolution Heterogeneous Face Hallucination.” 2022.
[BibTeX] [Link]@inproceedings{247839270, title = {Supplementary File: Escaping Data Scarcity for High-Resolution Heterogeneous Face Hallucination}, author = {{Yiqun Mei} and {Pengfei Guo} and {Vishal M. Patel}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/33e69c1a173789d721185c07510af20013a509bf}, } - Malsha V. Perera, Nithin Gopalakrishnan Nair, W. G. C. Bandara, and Vishal M. Patel, “SAR Despeckling Using a Denoising Diffusion Probabilistic Model,” in IEEE Geoscience and Remote Sensing Letters, 2022.
[BibTeX] [Link]@inproceedings{249538421, title = {SAR Despeckling Using a Denoising Diffusion Probabilistic Model}, author = {{Malsha V. Perera} and {Nithin Gopalakrishnan Nair} and {W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {6}, booktitle = {IEEE Geoscience and Remote Sensing Letters}, url = {https://www.semanticscholar.org/paper/946e217f0e734561ac0acd8b58063ca882c963df}, } - H. Kayser, H. Hermansky, and B. Meyer, “Spatial speech detection for binaural hearing aids using deep phoneme classifiers,” in Acta acustica. European Acoustics Association, 2022.
[BibTeX] [Link]@inproceedings{250105551, title = {Spatial speech detection for binaural hearing aids using deep phoneme classifiers}, author = {{H. Kayser} and {H. Hermansky} and {B. Meyer}}, year = 2022, month = {6}, booktitle = {Acta acustica. European Acoustics Association}, url = {https://www.semanticscholar.org/paper/5bf8888705bfa1cdbf08784606d5ebf6e6a0e2f8}, } - Yu Zeng, Zhe Lin, and Vishal M. Patel, “Shape-guided Object Inpainting,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248228101, title = {Shape-guided Object Inpainting}, author = {{Yu Zeng} and {Zhe Lin} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/69286603f2dd6037634921e1247543e30fe1756d}, } - W. G. C. Bandara and Vishal M. Patel, “Revisiting Consistency Regularization for Semi-supervised Change Detection in Remote Sensing Images,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{248227361, title = {Revisiting Consistency Regularization for Semi-supervised Change Detection in Remote Sensing Images}, author = {{W. G. C. Bandara} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3e0fda3d87152ddafdc8357765036c52a8f477d7}, } - Emma Bigelow, S. Saria, B. Piening, B. Curti, A. Dowdell, R. Weerasinghe, C. Bifulco, W. Urba, N. Finkelstein, E. Fertig, A. Baras, N. Zaidi, E. Jaffee, and M. Yarchoan, “A Random Forest Genomic Classifier for Tumor Agnostic Prediction of Response to Anti-PD1 Immunotherapy,” in Cancer Informatics, 2022.
[BibTeX] [Link]@inproceedings{253833423, title = {A Random Forest Genomic Classifier for Tumor Agnostic Prediction of Response to Anti-PD1 Immunotherapy}, author = {{Emma Bigelow} and {S. Saria} and {B. Piening} and {B. Curti} and {A. Dowdell} and {R. Weerasinghe} and {C. Bifulco} and {W. Urba} and {N. Finkelstein} and {E. Fertig} and {A. Baras} and {N. Zaidi} and {E. Jaffee} and {M. Yarchoan}}, year = 2022, month = {1}, booktitle = {Cancer Informatics}, url = {https://www.semanticscholar.org/paper/a407bd6bae19371a8d3c92da0981aaf1e80b382e}, } - Qihang Yu, Huiyu Wang, Siyuan Qiao, Maxwell D. Collins, Yukun Zhu, Hartwig Adam, A. Yuille, and Liang-Chieh Chen, “kMaX-DeepLab: k-means Mask Transformer.” 2022.
[BibTeX] [Link]@inproceedings{259692914, title = {kMaX-DeepLab: k-means Mask Transformer}, author = {{Qihang Yu} and {Huiyu Wang} and {Siyuan Qiao} and {Maxwell D. Collins} and {Yukun Zhu} and {Hartwig Adam} and {A. Yuille} and {Liang-Chieh Chen}}, year = 2022, month = {7}, booktitle = {}, url = {https://www.semanticscholar.org/paper/5f3c2a31fc84d13a72008f70106163bd92f2f9d0}, } - B. Vasey, M. Nagendran, Bruce Campbell, D. Clifton, Gary S. Collins, Spiros C. Denaxas, A. Denniston, L. Faes, B. Geerts, Mudathir Ibrahim, Xiaoxuan Liu, B. Mateen, P. Mathur, M. Mccradden, L. Morgan, Johan Ordish, Campbell Rogers, S. Saria, D. Ting, P. Watkinson, W. Weber, P. Wheatstone, and P. McCulloch, “Author Correction: Regenerative and restorative medicine for eye disease,” in Nature Network Boston, 2022.
[BibTeX] [Link]@inproceedings{251539930, title = {Author Correction: Regenerative and restorative medicine for eye disease}, author = {{B. Vasey} and {M. Nagendran} and {Bruce Campbell} and {D. Clifton} and {Gary S. Collins} and {Spiros C. Denaxas} and {A. Denniston} and {L. Faes} and {B. Geerts} and {Mudathir Ibrahim} and {Xiaoxuan Liu} and {B. Mateen} and {P. Mathur} and {M. Mccradden} and {L. Morgan} and {Johan Ordish} and {Campbell Rogers} and {S. Saria} and {D. Ting} and {P. Watkinson} and {W. Weber} and {P. Wheatstone} and {P. McCulloch}}, year = 2022, month = {8}, booktitle = {Nature Network Boston}, url = {https://www.semanticscholar.org/paper/a22215acadb4ad4ec04624025021023acf7261d6}, } - Sonal Joshi, Saurabh Kataria, Yiwen Shao, Piotr Żelasko, J. Villalba, S. Khudanpur, and N. Dehak, “Defense against Adversarial Attacks on Hybrid Speech Recognition System using Adversarial Fine-tuning with Denoiser,” in Interspeech, 2022.
[BibTeX] [Link]@inproceedings{252346818, title = {Defense against Adversarial Attacks on Hybrid Speech Recognition System using Adversarial Fine-tuning with Denoiser}, author = {{Sonal Joshi} and {Saurabh Kataria} and {Yiwen Shao} and {Piotr Żelasko} and {J. Villalba} and {S. Khudanpur} and {N. Dehak}}, year = 2022, month = {9}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/b8c3c97f239a1048b460d659a14110cc7f7a499e}, } - J. Villalba, B. J. Borgstrom, Saurabh Kataria, Magdalena Rybicka, C. Castillo, Jaejin Cho, Leibny Paola García-Perera, P. Torres-Carrasquillo, and N. Dehak, “Advances in Cross-Lingual and Cross-Source Audio-Visual Speaker Recognition: The JHU-MIT System for NIST SRE21,” in The Speaker and Language Recognition Workshop, 2022.
[BibTeX] [Link]@inproceedings{249827199, title = {Advances in Cross-Lingual and Cross-Source Audio-Visual Speaker Recognition: The JHU-MIT System for NIST SRE21}, author = {{J. Villalba} and {B. J. Borgstrom} and {Saurabh Kataria} and {Magdalena Rybicka} and {C. Castillo} and {Jaejin Cho} and {Leibny Paola García-Perera} and {P. Torres-Carrasquillo} and {N. Dehak}}, year = 2022, month = {6}, booktitle = {The Speaker and Language Recognition Workshop}, url = {https://www.semanticscholar.org/paper/9d9b5b782cbaf98bfb198b120c343d813c99ecf5}, } - Jaejin Cho, J. Villalba, L. Moro-Velázquez, and N. Dehak, “Non-Contrastive Self-Supervised Learning for Utterance-Level Information Extraction From Speech,” in IEEE Journal on Selected Topics in Signal Processing, 2022.
[BibTeX] [Link]@inproceedings{251462729, title = {Non-Contrastive Self-Supervised Learning for Utterance-Level Information Extraction From Speech}, author = {{Jaejin Cho} and {J. Villalba} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2022, month = {8}, booktitle = {IEEE Journal on Selected Topics in Signal Processing}, url = {https://www.semanticscholar.org/paper/7504aeee4c344c4cf9c6fc071dcc4b4b34d124cc}, } - Nils Holzenberger, Yunmo Chen, and Benjamin Van Durme, “Asking the Right Questions in Low Resource Template Extraction,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249062873, title = {Asking the Right Questions in Low Resource Template Extraction}, author = {{Nils Holzenberger} and {Yunmo Chen} and {Benjamin Van Durme}}, year = 2022, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/196b71b4e8465dd632954cf499f0467754cbd9d4}, } - Harminder Singh, R. Sharma, and R. Arora, “A Novel Dual-band filtenna for 2.4 and 5.8 GHz Wireless Local Area for Network Applications,” in 2022 Interdisciplinary Research in Technology and Management (IRTM), 2022.
[BibTeX] [Link]@inproceedings{249795778, title = {A Novel Dual-band filtenna for 2.4 and 5.8 GHz Wireless Local Area for Network Applications}, author = {{Harminder Singh} and {R. Sharma} and {R. Arora}}, year = 2022, month = {2}, booktitle = {2022 Interdisciplinary Research in Technology and Management (IRTM)}, url = {https://www.semanticscholar.org/paper/a773c6edcc796c34a4cd477d6a39043cab45d037}, } - Suraj Nair, Eugene Yang, Dawn J Lawrie, Kevin Duh, Paul McNamee, Kenton Murray, J. Mayfield, and Douglas W. Oard, “Transfer Learning Approaches for Building Cross-Language Dense Retrieval Models,” in European Conference on Information Retrieval, 2022.
[BibTeX] [Link]@inproceedings{246210468, title = {Transfer Learning Approaches for Building Cross-Language Dense Retrieval Models}, author = {{Suraj Nair} and {Eugene Yang} and {Dawn J Lawrie} and {Kevin Duh} and {Paul McNamee} and {Kenton Murray} and {J. Mayfield} and {Douglas W. Oard}}, year = 2022, month = {1}, booktitle = {European Conference on Information Retrieval}, url = {https://www.semanticscholar.org/paper/d1ccffb8eb1b7a99cd586723074b82fa5399bdd2}, } - Christos Sapsanis, M. Sophocleous, A. Andreou, and J. Georgiou, “Trade-Offs in Sensor Systems Design: A Tutorial,” in IEEE Sensors Journal, 2022.
[BibTeX] [Link]@inproceedings{246805990, title = {Trade-Offs in Sensor Systems Design: A Tutorial}, author = {{Christos Sapsanis} and {M. Sophocleous} and {A. Andreou} and {J. Georgiou}}, year = 2022, month = {6}, booktitle = {IEEE Sensors Journal}, url = {https://www.semanticscholar.org/paper/07cfa0c80e6ef73a2aa5fab377c2f698ed476341}, } - Aimon Rahman, Jeya Maria Jose Valanarasu, I. Hacihaliloglu, and Vishal M. Patel, “Simultaneous Bone and Shadow Segmentation Network using Task Correspondence Consistency,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2022.
[BibTeX] [Link]@inproceedings{249889093, title = {Simultaneous Bone and Shadow Segmentation Network using Task Correspondence Consistency}, author = {{Aimon Rahman} and {Jeya Maria Jose Valanarasu} and {I. Hacihaliloglu} and {Vishal M. Patel}}, year = 2022, month = {6}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/706ae2328d0207f956b7fd644b1bb64b130950e5}, } - Pengfei Guo, Yiqun Mei, Jinyuan Zhou, Shanshan Jiang, and Vishal M. Patel, “ReconFormer: Accelerated MRI Reconstruction Using Recurrent Transformer,” in IEEE Transactions on Medical Imaging, 2022.
[BibTeX] [Link]@inproceedings{246240294, title = {ReconFormer: Accelerated MRI Reconstruction Using Recurrent Transformer}, author = {{Pengfei Guo} and {Yiqun Mei} and {Jinyuan Zhou} and {Shanshan Jiang} and {Vishal M. Patel}}, year = 2022, month = {1}, booktitle = {IEEE Transactions on Medical Imaging}, url = {https://www.semanticscholar.org/paper/a54467e6d5df7727c1c0338dd79449dcc78c406e}, } - A. Kala, E. McCollum, and Mounya Elhilali, “Implications of clinical variability on computer-aided lung auscultation classification,” in Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2022.
[BibTeX] [Link]@inproceedings{252165718, title = {Implications of clinical variability on computer-aided lung auscultation classification}, author = {{A. Kala} and {E. McCollum} and {Mounya Elhilali}}, year = 2022, month = {7}, booktitle = {Annual International Conference of the IEEE Engineering in Medicine and Biology Society}, url = {https://www.semanticscholar.org/paper/f97aa46f0602e85f4254933ad709f8fd1a4ab35f}, } - Shota Horiguchi, Shinji Watanabe, Leibny Paola García-Perera, Yuki Takashima, and Y. Kawaguchi, “Online Neural Diarization of Unlimited Numbers of Speakers Using Global and Local Attractors,” in IEEE/ACM Transactions on Audio Speech and Language Processing, 2022.
[BibTeX] [Link]@inproceedings{249394670, title = {Online Neural Diarization of Unlimited Numbers of Speakers Using Global and Local Attractors}, author = {{Shota Horiguchi} and {Shinji Watanabe} and {Leibny Paola García-Perera} and {Yuki Takashima} and {Y. Kawaguchi}}, year = 2022, month = {6}, booktitle = {IEEE/ACM Transactions on Audio Speech and Language Processing}, url = {https://www.semanticscholar.org/paper/872c99ead3cc2644fbabd7dab37b82d233cc81cb}, } - Drew Prinster, Anqi Liu, and S. Saria, “JAWS: Auditing Predictive Uncertainty Under Covariate Shift,” in Neural Information Processing Systems, 2022.
[BibTeX] [Link]@inproceedings{254017908, title = {JAWS: Auditing Predictive Uncertainty Under Covariate Shift}, author = {{Drew Prinster} and {Anqi Liu} and {S. Saria}}, year = 2022, month = {7}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/4fb13897dad166844ca020e3cef1563b8dc81775}, } - Boyuan Zheng, Patrick Xia, M. Yarmohammadi, and Benjamin Van Durme, “Multilingual Coreference Resolution in Multiparty Dialogue,” in Transactions of the Association for Computational Linguistics, 2022.
[BibTeX] [Link]@inproceedings{251253396, title = {Multilingual Coreference Resolution in Multiparty Dialogue}, author = {{Boyuan Zheng} and {Patrick Xia} and {M. Yarmohammadi} and {Benjamin Van Durme}}, year = 2022, month = {8}, booktitle = {Transactions of the Association for Computational Linguistics}, url = {https://www.semanticscholar.org/paper/ced26dee12fe5fa2f666bcc2ba5b0a1969240887}, } - Nithin Gopalakrishnan Nair, Kangfu Mei, and Vishal M. Patel, “AT-DDPM: Restoring Faces Degraded by Atmospheric Turbulence Using Denoising Diffusion Probabilistic Models,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{251765371, title = {AT-DDPM: Restoring Faces Degraded by Atmospheric Turbulence Using Denoising Diffusion Probabilistic Models}, author = {{Nithin Gopalakrishnan Nair} and {Kangfu Mei} and {Vishal M. Patel}}, year = 2022, month = {8}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/dad4a46e1fe0e8317bd6734ffdf5609d1f577559}, } - Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao, Abu Awal Md Shoeb, Abubakar Abid, Adam Fisch, Adam R. Brown, Adam Santoro, Aditya Gupta, Adrià Garriga-Alonso, Agnieszka Kluska, Aitor Lewkowycz, Akshat Agarwal, Alethea Power, Alex Ray, Alex Warstadt, Alexander W. Kocurek, Ali Safaya, Ali Tazarv, Alice Xiang, Alicia Parrish, Allen Nie, A. Hussain, Amanda Askell, Amanda Dsouza, A. Rahane, Anantharaman S. Iyer, Anders Andreassen, Andrea Santilli, Andreas Stuhlmuller, Andrew M. Dai, Andrew D. La, Andrew Kyle Lampinen, Andy Zou, Angela Jiang, Angelica Chen, Anh Vuong, Animesh Gupta, Anna Gottardi, Antonio Norelli, Anu Venkatesh, Arash Gholamidavoodi, Arfa Tabassum, Arul Menezes, Arun Kirubarajan, A. Mullokandov, Ashish Sabharwal, Austin Herrick, Avia Efrat, Aykut Erdem, Ayla Karakacs, B. R. Roberts, B. S. Loe, Barret Zoph, Bartlomiej Bojanowski, Batuhan Ozyurt, Behnam Hedayatnia, Behnam Neyshabur, Benjamin Inden, Benno Stein, Berk Ekmekci, Bill Yuchen Lin, B. Howald, Cameron Diao, Cameron Dour, Catherine Stinson, Cedrick Argueta, C’esar Ferri Ram’irez, Chandan Singh, Charles Rathkopf, Chenlin Meng, Chitta Baral, Chiyu Wu, Chris Callison-Burch, Chris Waites, Christian Voigt, Christopher D. Manning, Christopher Potts, Cindy Ramirez, Clara Rivera, Clemencia Siro, Colin Raffel, Courtney Ashcraft, Cristina Garbacea, Damien Sileo, Daniel H Garrette, Dan Hendrycks, D. Kilman, D. Roth, Daniel Freeman, Daniel Khashabi, Daniel Levy, D. Gonz’alez, Danny Hernandez, Danqi Chen, Daphne Ippolito, D. Gilboa, David Dohan, D. Drakard, David Jurgens, Debajyoti Datta, Deep Ganguli, Denis Emelin, D. Kleyko, Deniz Yuret, Derek Chen, Derek Tam, D. Hupkes, Diganta Misra, Dilyar Buzan, Dimitri Coelho Mollo, Diyi Yang, Dong-Ho Lee, Ekaterina Shutova, E. D. Cubuk, Elad Segal, Eleanor Hagerman, Elizabeth Barnes, E. Donoway, Ellie Pavlick, E. Rodolà, E. Lam, Eric Chu, Eric Tang, Erkut Erdem, Ernie Chang, Ethan A. Chi, Ethan Dyer, E. Jerzak, Ethan Kim, Eunice Engefu Manyasi, Evgenii Zheltonozhskii, Fan Xia, F. Siar, Fernando Mart’inez-Plumed, Francesca Happ’e, François Chollet, Frieda Rong, Gaurav Mishra, Genta Indra Winata, Gerard de Melo, Germán Kruszewski, Giambattista Parascandolo, G. Mariani, Gloria Wang, Gonzalo Jaimovitch-L’opez, Gregor Betz, Guy Gur-Ari, Hana Galijasevic, H. Kim, Hannah Rashkin, Hannaneh Hajishirzi, Harsh Mehta, H. Bogar, Henry Shevlin, Hinrich Schütze, Hiromu Yakura, Hongming Zhang, Hubert Wong, I. Ng, Isaac Noble, Jaap Jumelet, Jack Geissinger, John Kernion, Jacob Hilton, Jaehoon Lee, J. Fisac, J. B. Simon, James Koppel, James Zheng, James Zou, Jan Koco’n, Jana Thompson, Jared Kaplan, Jarema Radom, Jascha Narain Sohl-Dickstein, Jason Phang, Jason Wei, J. Yosinski, Jekaterina Novikova, Jelle Bosscher, Jenni Marsh, Jeremy Kim, Jeroen Taal, Jesse Engel, Jesujoba Oluwadara Alabi, Jiacheng Xu, Jiaming Song, Jillian Tang, Jane W Waweru, John Burden, John Miller, John U. Balis, Jonathan Berant, Jorg Frohberg, Jos Rozen, J. Hernández-Orallo, Joseph Boudeman, Joseph Jones, J. Tenenbaum, Joshua S. Rule, Joyce Chua, Kamil Kanclerz, Karen Livescu, K. Krauth, Karthik Gopalakrishnan, Katerina Ignatyeva, K. Markert, Kaustubh D. Dhole, Kevin Gimpel, Kevin Omondi, K. Mathewson, Kristen Chiafullo, Ksenia Shkaruta, K. Shridhar, Kyle McDonell, Kyle Richardson, Laria Reynolds, Leo Gao, Li Zhang, Liam Dugan, Lianhui Qin, Lidia Contreras-Ochando, Louis-Philippe Morency, Luca Moschella, Luca Lam, Lucy Noble, Ludwig Schmidt, Luheng He, Luis Oliveros Col’on, Luke Metz, Lutfi Kerem cSenel, Maarten Bosma, Maarten Sap, Maartje ter Hoeve, Madotto Andrea, Maheen Farooqi, Manaal Faruqui, Mantas Mazeika, Marco Baturan, M. Marelli, Marco Maru, M. Quintana, Marie Tolkiehn, Mario Giulianelli, Martha Lewis, Martin Potthast, Matthew L. Leavitt, Matthias Hagen, M. Schubert, Medina Baitemirova, M. Arnaud, M. McElrath, Michael A. Yee, Michael Cohen, Mi Gu, Michael Ivanitskiy, Michael Starritt, M. Strube, Michal Swkedrowski, Michele Bevilacqua, Michihiro Yasunaga, Mihir Kale, Mike Cain, Mimee Xu, Mirac Suzgun, Monica Tiwari, Mohit Bansal, Moin Aminnaseri, Mor Geva, Mozhdeh Gheini, T. MukundVarma, Nanyun Peng, Nathan Chi, Nayeon Lee, Neta Gur-Ari Krakover, Nicholas Cameron, Nicholas S. Roberts, Nicholas Doiron, Nikita Nangia, Niklas Deckers, Niklas Muennighoff, N. Keskar, Niveditha Iyer, Noah Constant, Noah Fiedel, Nuan Wen, Oliver Zhang, Omar Agha, Omar Elbaghdadi, Omer Levy, Owain Evans, Pablo Antonio Moreno Casares, P. Doshi, Pascale Fung, P. Liang, Paul Vicol, Pegah Alipoormolabashi, Peiyuan Liao, Percy Liang, Peter Chang, P. Eckersley, Phu Mon Htut, Pi-Bei Hwang, P. Milkowski, P. Patil, Pouya Pezeshkpour, P. Oli, Q. Mei, Qing Lyu, Qinlang Chen, Rabin Banjade, R. Rudolph, Raefer Gabriel, Rahel Habacker, R. Delgado, Raphael Milliere, Rhythm Garg, Richard Barnes, R. Saurous, Riku Arakawa, Robbe Raymaekers, R. Frank, Rohan Sikand, Roman Novak, Roman Sitelew, Ronan Le Bras, Rosanne Liu, Rowan Jacobs, Rui Zhang, R. Salakhutdinov, Ryan Chi, Ryan Lee, Ryan Stovall, Ryan Teehan, Rylan Yang, Sahib J Singh, Saif M. Mohammad, Sajant Anand, Sam Dillavou, Sam Shleifer, Sam Wiseman, Samuel Gruetter, Sam Bowman, S. Schoenholz, Sanghyun Han, Sanjeev Kwatra, Sarah A. Rous, Sarik Ghazarian, Sayan Ghosh, S. Casey, Sebastian Bischoff, Sebastian Gehrmann, Sebastian Schuster, Sepideh Sadeghi, Shadi S. Hamdan, Sharon Zhou, Shashank Srivastava, Sherry Shi, S. Singh, Shima Asaadi, S. Gu, Shubh Pachchigar, Shubham Toshniwal, Shyam Upadhyay, Shyamolima Debnath, Siamak Shakeri, Simon Thormeyer, S. Melzi, Siva Reddy, S. Makini, Soo-hwan Lee, Spencer Bradley Torene, Sriharsha Hatwar, S. Dehaene, Stefan Divic, Stefano Ermon, Stella Biderman, Stephanie C. Lin, S. Prasad, S. Piantadosi, S. Shieber, Summer Misherghi, S. Kiritchenko, Swaroop Mishra, Tal Linzen, Tal Schuster, Tao Li, Tao Yu, Tariq A. Ali, Tatsuo Hashimoto, Te-Lin Wu, T. Desbordes, Theodore Rothschild, Thomas Phan, Tianle Wang, Tiberius Nkinyili, Timo Schick, T. Kornev, Timothy Telleen-Lawton, T. Tunduny, Tobias Gerstenberg, T. Chang, Trishala Neeraj, Tushar Khot, T. Shultz, Uri Shaham, Vedant Misra, Vera Demberg, Victoria Nyamai, Vikas Raunak, V. Ramasesh, Vinay Uday Prabhu, Vishakh Padmakumar, Vivek Srikumar, W. Fedus, W. Saunders, William Zhang, W. Vossen, Xiang Ren, Xiaoyu Tong, Xinyi Wu, Xudong Shen, Yadollah Yaghoobzadeh, Yair Lakretz, Yang Song, Yasaman Bahri, Y. Choi, Yichi Yang, Sophie Hao, Yifu Chen, Yonatan Belinkov, Yu Hou, Yu Hou, Yuntao Bai, Zachary Seid, Zhao Xinran, Zhuoye Zhao, Z. Wang, Zijie J. Wang, Zirui Wang, Ziyi Wu, Sahib Singh, and Uri Shaham, “Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249538544, title = {Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models}, author = {{Aarohi Srivastava} and {Abhinav Rastogi} and {Abhishek Rao} and {Abu Awal Md Shoeb} and {Abubakar Abid} and {Adam Fisch} and {Adam R. Brown} and {Adam Santoro} and {Aditya Gupta} and {Adrià Garriga-Alonso} and {Agnieszka Kluska} and {Aitor Lewkowycz} and {Akshat Agarwal} and {Alethea Power} and {Alex Ray} and {Alex Warstadt} and {Alexander W. Kocurek} and {Ali Safaya} and {Ali Tazarv} and {Alice Xiang} and {Alicia Parrish} and {Allen Nie} and {A. Hussain} and {Amanda Askell} and {Amanda Dsouza} and {A. Rahane} and {Anantharaman S. Iyer} and {Anders Andreassen} and {Andrea Santilli} and {Andreas Stuhlmuller} and {Andrew M. Dai} and {Andrew D. La} and {Andrew Kyle Lampinen} and {Andy Zou} and {Angela Jiang} and {Angelica Chen} and {Anh Vuong} and {Animesh Gupta} and {Anna Gottardi} and {Antonio Norelli} and {Anu Venkatesh} and {Arash Gholamidavoodi} and {Arfa Tabassum} and {Arul Menezes} and {Arun Kirubarajan} and {A. Mullokandov} and {Ashish Sabharwal} and {Austin Herrick} and {Avia Efrat} and {Aykut Erdem} and {Ayla Karakacs} and {B. R. Roberts} and {B. S. Loe} and {Barret Zoph} and {Bartlomiej Bojanowski} and {Batuhan Ozyurt} and {Behnam Hedayatnia} and {Behnam Neyshabur} and {Benjamin Inden} and {Benno Stein} and {Berk Ekmekci} and {Bill Yuchen Lin} and {B. Howald} and {Cameron Diao} and {Cameron Dour} and {Catherine Stinson} and {Cedrick Argueta} and {C'esar Ferri Ram'irez} and {Chandan Singh} and {Charles Rathkopf} and {Chenlin Meng} and {Chitta Baral} and {Chiyu Wu} and {Chris Callison-Burch} and {Chris Waites} and {Christian Voigt} and {Christopher D. Manning} and {Christopher Potts} and {Cindy Ramirez} and {Clara Rivera} and {Clemencia Siro} and {Colin Raffel} and {Courtney Ashcraft} and {Cristina Garbacea} and {Damien Sileo} and {Daniel H Garrette} and {Dan Hendrycks} and {D. Kilman} and {D. Roth} and {Daniel Freeman} and {Daniel Khashabi} and {Daniel Levy} and {D. Gonz'alez} and {Danny Hernandez} and {Danqi Chen} and {Daphne Ippolito} and {D. Gilboa} and {David Dohan} and {D. Drakard} and {David Jurgens} and {Debajyoti Datta} and {Deep Ganguli} and {Denis Emelin} and {D. Kleyko} and {Deniz Yuret} and {Derek Chen} and {Derek Tam} and {D. Hupkes} and {Diganta Misra} and {Dilyar Buzan} and {Dimitri Coelho Mollo} and {Diyi Yang} and {Dong-Ho Lee} and {Ekaterina Shutova} and {E. D. Cubuk} and {Elad Segal} and {Eleanor Hagerman} and {Elizabeth Barnes} and {E. Donoway} and {Ellie Pavlick} and {E. Rodolà} and {E. Lam} and {Eric Chu} and {Eric Tang} and {Erkut Erdem} and {Ernie Chang} and {Ethan A. Chi} and {Ethan Dyer} and {E. Jerzak} and {Ethan Kim} and {Eunice Engefu Manyasi} and {Evgenii Zheltonozhskii} and {Fan Xia} and {F. Siar} and {Fernando Mart'inez-Plumed} and {Francesca Happ'e} and {François Chollet} and {Frieda Rong} and {Gaurav Mishra} and {Genta Indra Winata} and {Gerard de Melo} and {Germán Kruszewski} and {Giambattista Parascandolo} and {G. Mariani} and {Gloria Wang} and {Gonzalo Jaimovitch-L'opez} and {Gregor Betz} and {Guy Gur-Ari} and {Hana Galijasevic} and {H. Kim} and {Hannah Rashkin} and {Hannaneh Hajishirzi} and {Harsh Mehta} and {H. Bogar} and {Henry Shevlin} and {Hinrich Schütze} and {Hiromu Yakura} and {Hongming Zhang} and {Hubert Wong} and {I. Ng} and {Isaac Noble} and {Jaap Jumelet} and {Jack Geissinger} and {John Kernion} and {Jacob Hilton} and {Jaehoon Lee} and {J. Fisac} and {J. B. Simon} and {James Koppel} and {James Zheng} and {James Zou} and {Jan Koco'n} and {Jana Thompson} and {Jared Kaplan} and {Jarema Radom} and {Jascha Narain Sohl-Dickstein} and {Jason Phang} and {Jason Wei} and {J. Yosinski} and {Jekaterina Novikova} and {Jelle Bosscher} and {Jenni Marsh} and {Jeremy Kim} and {Jeroen Taal} and {Jesse Engel} and {Jesujoba Oluwadara Alabi} and {Jiacheng Xu} and {Jiaming Song} and {Jillian Tang} and {Jane W Waweru} and {John Burden} and {John Miller} and {John U. Balis} and {Jonathan Berant} and {Jorg Frohberg} and {Jos Rozen} and {J. Hernández-Orallo} and {Joseph Boudeman} and {Joseph Jones} and {J. Tenenbaum} and {Joshua S. Rule} and {Joyce Chua} and {Kamil Kanclerz} and {Karen Livescu} and {K. Krauth} and {Karthik Gopalakrishnan} and {Katerina Ignatyeva} and {K. Markert} and {Kaustubh D. Dhole} and {Kevin Gimpel} and {Kevin Omondi} and {K. Mathewson} and {Kristen Chiafullo} and {Ksenia Shkaruta} and {K. Shridhar} and {Kyle McDonell} and {Kyle Richardson} and {Laria Reynolds} and {Leo Gao} and {Li Zhang} and {Liam Dugan} and {Lianhui Qin} and {Lidia Contreras-Ochando} and {Louis-Philippe Morency} and {Luca Moschella} and {Luca Lam} and {Lucy Noble} and {Ludwig Schmidt} and {Luheng He} and {Luis Oliveros Col'on} and {Luke Metz} and {Lutfi Kerem cSenel} and {Maarten Bosma} and {Maarten Sap} and {Maartje ter Hoeve} and {Madotto Andrea} and {Maheen Farooqi} and {Manaal Faruqui} and {Mantas Mazeika} and {Marco Baturan} and {M. Marelli} and {Marco Maru} and {M. Quintana} and {Marie Tolkiehn} and {Mario Giulianelli} and {Martha Lewis} and {Martin Potthast} and {Matthew L. Leavitt} and {Matthias Hagen} and {M. Schubert} and {Medina Baitemirova} and {M. Arnaud} and {M. McElrath} and {Michael A. Yee} and {Michael Cohen} and {Mi Gu} and {Michael Ivanitskiy} and {Michael Starritt} and {M. Strube} and {Michal Swkedrowski} and {Michele Bevilacqua} and {Michihiro Yasunaga} and {Mihir Kale} and {Mike Cain} and {Mimee Xu} and {Mirac Suzgun} and {Monica Tiwari} and {Mohit Bansal} and {Moin Aminnaseri} and {Mor Geva} and {Mozhdeh Gheini} and {T. MukundVarma} and {Nanyun Peng} and {Nathan Chi} and {Nayeon Lee} and {Neta Gur-Ari Krakover} and {Nicholas Cameron} and {Nicholas S. Roberts} and {Nicholas Doiron} and {Nikita Nangia} and {Niklas Deckers} and {Niklas Muennighoff} and {N. Keskar} and {Niveditha Iyer} and {Noah Constant} and {Noah Fiedel} and {Nuan Wen} and {Oliver Zhang} and {Omar Agha} and {Omar Elbaghdadi} and {Omer Levy} and {Owain Evans} and {Pablo Antonio Moreno Casares} and {P. Doshi} and {Pascale Fung} and {P. Liang} and {Paul Vicol} and {Pegah Alipoormolabashi} and {Peiyuan Liao} and {Percy Liang} and {Peter Chang} and {P. Eckersley} and {Phu Mon Htut} and {Pi-Bei Hwang} and {P. Milkowski} and {P. Patil} and {Pouya Pezeshkpour} and {P. Oli} and {Q. Mei} and {Qing Lyu} and {Qinlang Chen} and {Rabin Banjade} and {R. Rudolph} and {Raefer Gabriel} and {Rahel Habacker} and {R. Delgado} and {Raphael Milliere} and {Rhythm Garg} and {Richard Barnes} and {R. Saurous} and {Riku Arakawa} and {Robbe Raymaekers} and {R. Frank} and {Rohan Sikand} and {Roman Novak} and {Roman Sitelew} and {Ronan Le Bras} and {Rosanne Liu} and {Rowan Jacobs} and {Rui Zhang} and {R. Salakhutdinov} and {Ryan Chi} and {Ryan Lee} and {Ryan Stovall} and {Ryan Teehan} and {Rylan Yang} and {Sahib J Singh} and {Saif M. Mohammad} and {Sajant Anand} and {Sam Dillavou} and {Sam Shleifer} and {Sam Wiseman} and {Samuel Gruetter} and {Sam Bowman} and {S. Schoenholz} and {Sanghyun Han} and {Sanjeev Kwatra} and {Sarah A. Rous} and {Sarik Ghazarian} and {Sayan Ghosh} and {S. Casey} and {Sebastian Bischoff} and {Sebastian Gehrmann} and {Sebastian Schuster} and {Sepideh Sadeghi} and {Shadi S. Hamdan} and {Sharon Zhou} and {Shashank Srivastava} and {Sherry Shi} and {S. Singh} and {Shima Asaadi} and {S. Gu} and {Shubh Pachchigar} and {Shubham Toshniwal} and {Shyam Upadhyay} and {Shyamolima Debnath} and {Siamak Shakeri} and {Simon Thormeyer} and {S. Melzi} and {Siva Reddy} and {S. Makini} and {Soo-hwan Lee} and {Spencer Bradley Torene} and {Sriharsha Hatwar} and {S. Dehaene} and {Stefan Divic} and {Stefano Ermon} and {Stella Biderman} and {Stephanie C. Lin} and {S. Prasad} and {S. Piantadosi} and {S. Shieber} and {Summer Misherghi} and {S. Kiritchenko} and {Swaroop Mishra} and {Tal Linzen} and {Tal Schuster} and {Tao Li} and {Tao Yu} and {Tariq A. Ali} and {Tatsuo Hashimoto} and {Te-Lin Wu} and {T. Desbordes} and {Theodore Rothschild} and {Thomas Phan} and {Tianle Wang} and {Tiberius Nkinyili} and {Timo Schick} and {T. Kornev} and {Timothy Telleen-Lawton} and {T. Tunduny} and {Tobias Gerstenberg} and {T. Chang} and {Trishala Neeraj} and {Tushar Khot} and {T. Shultz} and {Uri Shaham} and {Vedant Misra} and {Vera Demberg} and {Victoria Nyamai} and {Vikas Raunak} and {V. Ramasesh} and {Vinay Uday Prabhu} and {Vishakh Padmakumar} and {Vivek Srikumar} and {W. Fedus} and {W. Saunders} and {William Zhang} and {W. Vossen} and {Xiang Ren} and {Xiaoyu Tong} and {Xinyi Wu} and {Xudong Shen} and {Yadollah Yaghoobzadeh} and {Yair Lakretz} and {Yang Song} and {Yasaman Bahri} and {Y. Choi} and {Yichi Yang} and {Sophie Hao} and {Yifu Chen} and {Yonatan Belinkov} and {Yu Hou} and {Yu Hou} and {Yuntao Bai} and {Zachary Seid} and {Zhao Xinran} and {Zhuoye Zhao} and {Z. Wang} and {Zijie J. Wang} and {Zirui Wang} and {Ziyi Wu} and {Sahib Singh} and {Uri Shaham}}, year = 2022, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/34503c0b6a615124eaf82cb0e4a1dab2866e8980}, } - Geoffrey M. Gray, L. Ahumada, Ayah Zirikly, Masoud Rouhizadeh, Tom M. Richards, and E. Hatef, “Application of Natural Language Processing to Identify Social Needs from The Electronic Health Record’s Free-Text Notes,” in American Medical Informatics Association Annual Symposium, 2022.
[BibTeX] [Link]@inproceedings{256665944, title = {Application of Natural Language Processing to Identify Social Needs from The Electronic Health Record's Free-Text Notes}, author = {{Geoffrey M. Gray} and {L. Ahumada} and {Ayah Zirikly} and {Masoud Rouhizadeh} and {Tom M. Richards} and {E. Hatef}}, year = 2022, booktitle = {American Medical Informatics Association Annual Symposium}, url = {https://www.semanticscholar.org/paper/9b579eeb9351a75c1c491f22f28ae36bdadded28}, } - J. Villalba, B. J. Borgstrom, Saurabh Kataria, Jaejin Cho, P. Torres-Carrasquillo, and N. Dehak, “Advances in Speaker Recognition for Multilingual Conversational Telephone Speech: The JHU-MIT System for NIST SRE20 CTS Challenge,” in The Speaker and Language Recognition Workshop, 2022.
[BibTeX] [Link]@inproceedings{249830266, title = {Advances in Speaker Recognition for Multilingual Conversational Telephone Speech: The JHU-MIT System for NIST SRE20 CTS Challenge}, author = {{J. Villalba} and {B. J. Borgstrom} and {Saurabh Kataria} and {Jaejin Cho} and {P. Torres-Carrasquillo} and {N. Dehak}}, year = 2022, month = {6}, booktitle = {The Speaker and Language Recognition Workshop}, url = {https://www.semanticscholar.org/paper/042e35459f6dfd8ad8be0dad72ae27f8e73cd4a8}, } - Jieru Mei, Yucheng Han, Yutong Bai, Yixiao Zhang, Yingwei Li, Xianhang Li, A. Yuille, and Cihang Xie, “Fast AdvProp,” in International Conference on Learning Representations, 2022.
[BibTeX] [Link]@inproceedings{248300043, title = {Fast AdvProp}, author = {{Jieru Mei} and {Yucheng Han} and {Yutong Bai} and {Yixiao Zhang} and {Yingwei Li} and {Xianhang Li} and {A. Yuille} and {Cihang Xie}}, year = 2022, month = {4}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/f2eaaf8afc89a86035fd7127305f2bb9d2169495}, } - A. Shelton, E. Davis, Cathryn S. Cortesa, Jonathan D. Jones, Gregory Hager, S. Khudanpur, and B. Landau, “Characterizing the Details of Spatial Construction: Cognitive Constraints and Variability,” in Cognitive Sciences, 2022.
[BibTeX] [Link]@inproceedings{246239741, title = {Characterizing the Details of Spatial Construction: Cognitive Constraints and Variability}, author = {{A. Shelton} and {E. Davis} and {Cathryn S. Cortesa} and {Jonathan D. Jones} and {Gregory Hager} and {S. Khudanpur} and {B. Landau}}, year = 2022, month = {1}, booktitle = {Cognitive Sciences}, url = {https://www.semanticscholar.org/paper/6482f52977f167c6db734f766b0b59e8c92d7e52}, } - Zengle Zhu, Mintong Kang, A. Yuille, and Zongwei Zhou, “Assembling Existing Labels from Public Datasets to Diagnose Novel Diseases: COVID-19 in Late 2019.” 2022.
[BibTeX] [Link]@inproceedings{259840482, title = {Assembling Existing Labels from Public Datasets to Diagnose Novel Diseases: COVID-19 in Late 2019}, author = {{Zengle Zhu} and {Mintong Kang} and {A. Yuille} and {Zongwei Zhou}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/5e9e11dbe87d01e44fc3a4e68d151f2a2809f261}, } - K. Henry, R. Adams, Cassandra Parent, Hossein Soleimani, A. Sridharan, Lauren Johnson, D. Hager, S. Cosgrove, Andrew Markowski, E. Klein, E. Chen, M. Saheed, Maureen Henley, S. Miranda, Katrina Houston, Robert C. Linton, Anushree R Ahluwalia, Albert W. Wu, and S. Saria, “Factors driving provider adoption of the TREWS machine learning-based early warning system and its effects on sepsis treatment timing,” in Nature Network Boston, 2022.
[BibTeX] [Link]@inproceedings{250953863, title = {Factors driving provider adoption of the TREWS machine learning-based early warning system and its effects on sepsis treatment timing}, author = {{K. Henry} and {R. Adams} and {Cassandra Parent} and {Hossein Soleimani} and {A. Sridharan} and {Lauren Johnson} and {D. Hager} and {S. Cosgrove} and {Andrew Markowski} and {E. Klein} and {E. Chen} and {M. Saheed} and {Maureen Henley} and {S. Miranda} and {Katrina Houston} and {Robert C. Linton} and {Anushree R Ahluwalia} and {Albert W. Wu} and {S. Saria}}, year = 2022, month = {7}, booktitle = {Nature Network Boston}, url = {https://www.semanticscholar.org/paper/4cf1afc1e27d26a77aca58d7a5ec7fe3d6b7ffad}, } - R. Yasarla and Vishal M. Patel, “CNN-Based Restoration of a Single Face Image Degraded by Atmospheric Turbulence,” in IEEE Transactions on Biometrics Behavior and Identity Science, 2022.
[BibTeX] [Link]@inproceedings{249929578, title = {CNN-Based Restoration of a Single Face Image Degraded by Atmospheric Turbulence}, author = {{R. Yasarla} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {IEEE Transactions on Biometrics Behavior and Identity Science}, url = {https://www.semanticscholar.org/paper/59f5937d4d7a81185a7a0501059c42cee271432f}, } - H. E. Echo Wang, M. Landers, R. Adams, Adarsh Subbaswamy, Hadi Kharrazi, D. Gaskin, and S. Saria, “A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models,” in J. Am. Medical Informatics Assoc., 2022.
[BibTeX] [Link]@inproceedings{248832471, title = {A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models}, author = {{H. E. Echo Wang} and {M. Landers} and {R. Adams} and {Adarsh Subbaswamy} and {Hadi Kharrazi} and {D. Gaskin} and {S. Saria}}, year = 2022, month = {5}, booktitle = {J. Am. Medical Informatics Assoc.}, url = {https://www.semanticscholar.org/paper/cdb65bc7700f365cf5ff152b6f3cb7434d9ad7e8}, } - Kangfu Mei, Vishal M. Patel, and Rui Huang, “Deep Semantic Statistics Matching (D2SM) Denoising Network,” in European Conference on Computer Vision, 2022.
[BibTeX] [Link]@inproceedings{250644264, title = {Deep Semantic Statistics Matching (D2SM) Denoising Network}, author = {{Kangfu Mei} and {Vishal M. Patel} and {Rui Huang}}, year = 2022, month = {7}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/19f83c24c56904754be700247b416cee704d5738}, } - Bruce Y Lee, J. Ordovás, E. J. Parks, Cheryl AM Anderson, A. Barabási, S. Clinton, K. Haye, V. Duffy, P. Franks, Elizabeth M Ginexi, K. Hammond, Erin C. Hanlon, Michael Hittle, Emily Ho, A. Horn, R. Isaacson, P. Mabry, Susan E. Malone, Corby K. Martin, J. Mattei, S. Meydani, Lorene M. Nelson, M. Neuhouser, N. Pronk, S. Saria, Frank Ajl Scheer, E. Segal, M. Sevick, T. Spector, Linda B Van Horn, K. Varady, V. S. Voruganti, and Marie F Martinez, “Microsoft Word-nqac237.docx.” 2022.
[BibTeX] [Link]@inproceedings{253253453, title = {Microsoft Word-nqac237.docx}, author = {{Bruce Y Lee} and {J. Ordovás} and {E. J. Parks} and {Cheryl AM Anderson} and {A. Barabási} and {S. Clinton} and {K. Haye} and {V. Duffy} and {P. Franks} and {Elizabeth M Ginexi} and {K. Hammond} and {Erin C. Hanlon} and {Michael Hittle} and {Emily Ho} and {A. Horn} and {R. Isaacson} and {P. Mabry} and {Susan E. Malone} and {Corby K. Martin} and {J. Mattei} and {S. Meydani} and {Lorene M. Nelson} and {M. Neuhouser} and {N. Pronk} and {S. Saria} and {Frank Ajl Scheer} and {E. Segal} and {M. Sevick} and {T. Spector} and {Linda B Van Horn} and {K. Varady} and {V. S. Voruganti} and {Marie F Martinez}}, year = 2022, booktitle = {}, url = {https://www.semanticscholar.org/paper/ae27ca3ffeb8273f258fb6a41a1cc4803adb716b}, } - W. G. C. Bandara, Nithin Gopalakrishnan Nair, and Vishal M. Patel, “DDPM-CD: Remote Sensing Change Detection using Denoising Diffusion Probabilistic Models,” in arXiv.org, 2022.
[BibTeX] [Link]@inproceedings{249953553, title = {DDPM-CD: Remote Sensing Change Detection using Denoising Diffusion Probabilistic Models}, author = {{W. G. C. Bandara} and {Nithin Gopalakrishnan Nair} and {Vishal M. Patel}}, year = 2022, month = {6}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/69a483159b03543e0a750776057218674287953b}, } - Yukun Feng, Feng Li, Ziang Song, Boyuan Zheng, and Philipp Koehn, “Learn To Remember: Transformer with Recurrent Memory for Document-Level Machine Translation,” in NAACL-HLT, 2022.
[BibTeX] [Link]@inproceedings{248505905, title = {Learn To Remember: Transformer with Recurrent Memory for Document-Level Machine Translation}, author = {{Yukun Feng} and {Feng Li} and {Ziang Song} and {Boyuan Zheng} and {Philipp Koehn}}, year = 2022, month = {5}, booktitle = {NAACL-HLT}, url = {https://www.semanticscholar.org/paper/4293121e2bef84aa8db5aab6634cfcd2d06947d4}, } - Jeya Maria Jose Valanarasu and Vishal M. Patel, “UNeXt: MLP-based Rapid Medical Image Segmentation Network,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2022.
[BibTeX] [Link]@inproceedings{247362702, title = {UNeXt: MLP-based Rapid Medical Image Segmentation Network}, author = {{Jeya Maria Jose Valanarasu} and {Vishal M. Patel}}, year = 2022, month = {3}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/ccb5a70f8a6f7b7fc923b9d4c18488b2837daa6f}, } - Nithin Gopalakrishnan Nair, Kangfu Mei, and Vishal M. Patel, “A Comparison of Different Atmospheric Turbulence Simulation Methods for Image Restoration,” in International Conference on Information Photonics, 2022.
[BibTeX] [Link]@inproceedings{248239720, title = {A Comparison of Different Atmospheric Turbulence Simulation Methods for Image Restoration}, author = {{Nithin Gopalakrishnan Nair} and {Kangfu Mei} and {Vishal M. Patel}}, year = 2022, month = {4}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/be3eb6827c645f176e204dffb5d740e5281dd67c}, } - Nathaniel Weir, Xingdi Yuan, Marc-Alexandre Côté, Matthew J. Hausknecht, R. Laroche, I. Momennejad, H. V. Seijen, and Benjamin Van Durme, “One-Shot Learning from a Demonstration with Hierarchical Latent Language,” in Adaptive Agents and Multi-Agent Systems, 2022.
[BibTeX] [Link]@inproceedings{247318673, title = {One-Shot Learning from a Demonstration with Hierarchical Latent Language}, author = {{Nathaniel Weir} and {Xingdi Yuan} and {Marc-Alexandre Côté} and {Matthew J. Hausknecht} and {R. Laroche} and {I. Momennejad} and {H. V. Seijen} and {Benjamin Van Durme}}, year = 2022, month = {3}, booktitle = {Adaptive Agents and Multi-Agent Systems}, url = {https://www.semanticscholar.org/paper/1d41a0ddda57caa6c8d268dd1703e4c9b35db18b}, } - Yuyin Zhou, D. Dreizin, Yan Wang, Fengze Liu, Wei-lei Shen, and A. Yuille, “External Attention Assisted Multi-Phase Splenic Vascular Injury Segmentation With Limited Data,” in IEEE Transactions on Medical Imaging, 2021.
[BibTeX] [Link]@inproceedings{245594119, title = {External Attention Assisted Multi-Phase Splenic Vascular Injury Segmentation With Limited Data}, author = {{Yuyin Zhou} and {D. Dreizin} and {Yan Wang} and {Fengze Liu} and {Wei-lei Shen} and {A. Yuille}}, year = 2021, month = {12}, booktitle = {IEEE Transactions on Medical Imaging}, url = {https://www.semanticscholar.org/paper/286f82f75ac6a7baa342217296d68eff30c07af6}, } - R. Adams, K. Henry, Hossein Soleimani, Nishi Rawat, M. Saheed, E. Chen, Albert W. Wu, and S. Saria, “1405: ASSESSING CLINICAL USE AND PERFORMANCE OF A MACHINE LEARNING SEPSIS ALERT FOR SEX AND RACIAL BIAS,” in Critical Care Medicine, 2021.
[BibTeX] [Link]@inproceedings{245295789, title = {1405: ASSESSING CLINICAL USE AND PERFORMANCE OF A MACHINE LEARNING SEPSIS ALERT FOR SEX AND RACIAL BIAS}, author = {{R. Adams} and {K. Henry} and {Hossein Soleimani} and {Nishi Rawat} and {M. Saheed} and {E. Chen} and {Albert W. Wu} and {S. Saria}}, year = 2021, month = {12}, booktitle = {Critical Care Medicine}, url = {https://www.semanticscholar.org/paper/b3c964ad654a01ccd25db4eb89129b7e8fb6bed1}, } - Chen Wei, Haoqi Fan, Saining Xie, Chaoxia Wu, A. Yuille, and Christoph Feichtenhofer, “Masked Feature Prediction for Self-Supervised Visual Pre-Training,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{245218767, title = {Masked Feature Prediction for Self-Supervised Visual Pre-Training}, author = {{Chen Wei} and {Haoqi Fan} and {Saining Xie} and {Chaoxia Wu} and {A. Yuille} and {Christoph Feichtenhofer}}, year = 2021, month = {12}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/008a428e049003fe768068a0f1fa1416af5c4982}, } - Kangfu Mei and Vishal M. Patel, “LTT-GAN: Looking Through Turbulence by Inverting GANs,” in IEEE Journal on Selected Topics in Signal Processing, 2021.
[BibTeX] [Link]@inproceedings{244909001, title = {LTT-GAN: Looking Through Turbulence by Inverting GANs}, author = {{Kangfu Mei} and {Vishal M. Patel}}, year = 2021, month = {12}, booktitle = {IEEE Journal on Selected Topics in Signal Processing}, url = {https://www.semanticscholar.org/paper/49b7c72a65263f7f0f1eaa4fd23ef3b73c013de1}, } - Chenglin Yang, Yilin Wang, Jianming Zhang, He Zhang, Zijun Wei, Zhe L. Lin, and A. Yuille, “Lite Vision Transformer with Enhanced Self-Attention,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{245353696, title = {Lite Vision Transformer with Enhanced Self-Attention}, author = {{Chenglin Yang} and {Yilin Wang} and {Jianming Zhang} and {He Zhang} and {Zijun Wei} and {Zhe L. Lin} and {A. Yuille}}, year = 2021, month = {12}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/72e81bc41ffae1d414836169107910025aaacb75}, } - Jingye Chen, Jieneng Chen, Zongwei Zhou, Bin Li, A. Yuille, and Yongyi Lu, “MT-TransUNet: Mediating Multi-Task Tokens in Transformers for Skin Lesion Segmentation and Classification,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{244896502, title = {MT-TransUNet: Mediating Multi-Task Tokens in Transformers for Skin Lesion Segmentation and Classification}, author = {{Jingye Chen} and {Jieneng Chen} and {Zongwei Zhou} and {Bin Li} and {A. Yuille} and {Yongyi Lu}}, year = 2021, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/aca5ae479a48aaee0b02cf07b55e909abee51ebb}, } - S. Saria, K. Henry, Hossein Soleimani, R. Adams, A. Zhan, Nishi Rawat, E. Chen, and Albert W. Wu, “1429: LEAD TIME AND ACCURACY OF TREWS, A MACHINE LEARNING-BASED SEPSIS ALERT,” in Critical Care Medicine, 2021.
[BibTeX] [Link]@inproceedings{245284251, title = {1429: LEAD TIME AND ACCURACY OF TREWS, A MACHINE LEARNING-BASED SEPSIS ALERT}, author = {{S. Saria} and {K. Henry} and {Hossein Soleimani} and {R. Adams} and {A. Zhan} and {Nishi Rawat} and {E. Chen} and {Albert W. Wu}}, year = 2021, month = {12}, booktitle = {Critical Care Medicine}, url = {https://www.semanticscholar.org/paper/66869b66e3beb408ffbccc97721678dc8c38963d}, } - Ju He, Shuo Yang, Shaokang Yang, Adam Kortylewski, Xiaoding Yuan, Jieneng Chen, Shuai Liu, Cheng Yang, and A. Yuille, “PartImageNet: A Large, High-Quality Dataset of Parts,” in European Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{244798652, title = {PartImageNet: A Large, High-Quality Dataset of Parts}, author = {{Ju He} and {Shuo Yang} and {Shaokang Yang} and {Adam Kortylewski} and {Xiaoding Yuan} and {Jieneng Chen} and {Shuai Liu} and {Cheng Yang} and {A. Yuille}}, year = 2021, month = {12}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/5c1dd63a45dc56009d1d499c8c2f4d7b9953a507}, } - Ryan T. Scott, E. Antonsen, L. Sanders, Jaden J. A. Hastings, Seung-min Park, Graham Mackintosh, R. Reynolds, A. Hoarfrost, A. Sawyer, C. Greene, Benjamin S. Glicksberg, C. Theriot, D. Berrios, Jack M. Miller, Joel Babdor, Richard Barker, S. Baranzini, Afshin Beheshti, S. Chalk, Guillermo M. Delgado-Aparicio, M. Haendel, Arif A. Hamid, P. Heller, Daniel Jamieson, K. Jarvis, John Kalantari, K. Khezeli, S. Komarova, M. Komorowski, Prachi Kothiyal, A. Mahabal, U. Manor, H. Martín, Christopher E. Mason, Mona Matar, G. Mias, J. Myers, Jr, Charlotte A. Nelson, Jonathan Oribello, P. Parsons-Wingerter, R. Prabhu, A. Qutub, J. Rask, Amanda M. Saravia-Butler, S. Saria, N. Singh, Frank Soboczenski, M. Snyder, Karthik Soman, D. V. Valen, K. Venkateswaran, L. Warren, Liz Worthey, Jason H. Yang, M. Zitnik, S. V. C. Kbr, Space Biosciences Division, N. R. Center, M. Field, Ca, USA., Department of Preventive Medicine, Center for Individualized Medicine, Baylor College of Medicine, Houston, Tx, Blue Marble Space Institute of Science, D. Physiology, Biophysics, Weill Cornell Medicine, N. York, Ny, D. Urology, D. Radiology, S. Medicine, Stanford, Bay Area Environmental Research Institute, Mortality ResearchConsulting, Inc., Universities Space Research Association, UC Space Health, D. Surgery, U. California, San Francisco, AI CenterforHealth, D. Biochemistry, Molecular Genetics, University of Maryland School of Medicine, Anschutz Medical Campus, Aurora, Co, Hasso Plattner Institute for Digital Health at Mount Sinai, Department of Genetics, Genomic Sciences, I. A. Sinai, Department of Preventive Medicine, C. Health, Utmb, Galveston, Human Health, Performance Directorate, NASAMarshall Space Flight Center, D. Microbiology, Immunology, Department of Otolaryngology, Head, N. Surgery, University of San Francisco, The Gilroy AstroBiology Research Group, The University of Wisconsin, Madison, Wi, Weill Institute for Neurosciences, D. Neurology, D. Chemistry, U. Florida, Jacksonville, Fl, D. Analytics, G. I. O. Technology, Lima, Perú, Department of Neuroscience, U. Minnesota, Minneapolis, Mn, Department of Materials Science, College of Materials Science, San Diego State University, San José, Biorelate, Manchester, United Kingdom., Center for Individualized Medicine, D. Surgery, D. Sciences, Mayo Clinic, Rochester, Faculty of Veterinary Medicine, Oral Health Sciences, M. University, Montreal., Quebec., Canada., Faculty of Veterinary Medicine, Cancer, I. -. London, London, SymbioSeq Llc, Ashburn, Va, Center for Data Driven Discovery, C. I. O. Technology., Pasadena, Waitt Advanced Biophotonics Center, Chan-Zuckerberg Imaging Scientist Fellow, Salk Institute for Biological Studies, La Jolla, Biological Systems, Engineering Division, Lawrence Berkeley National Lab., Berkeley, Doe Agile BioFoundry, Emeryville, Joint BioEnergy Institute, Human Research Program Cross-cutting Computational Model Project, N. R. Center, Cleveland, Oh, Institute for Computational Science, Engineering, M. Biology, M. University, E. Lansing., Mi, Low Exploration Gravity Technology, AI Matrix Consortium, Department of Biomedical Engineering, U. Texas, S. Antonio, UT Health Sciences, Office of the Director, Logyx, Computer Science, Statistics, H. Policy, J. University, Baltimore., Md., Ml, Ai, H. Lab, B. Health, Biotechnology, Planetary Protection Group, J. P. Laboratory, Sphes, Medical Faculty, King’s College London, S. Medicine, Department of Biology, Iss National Laboratory, Center for Space, Melbourne, Uab Center for Computational Biology, Data Science, U. Alabama, Birmingham, Al, Center for Emerging, Re-Emerging Pathogens, Biochemistry, Rutgers New Jersey Medical School, Newark, Nj, Department of Biomedical Informatics, Harvard Medical School, Harvard Data Science, Broad Institute of Mit, Harvard, Harvard University, Boston, and Ma., “Beyond Low Earth Orbit: Biomonitoring, Artificial Intelligence, and Precision Space Health,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{245425181, title = {Beyond Low Earth Orbit: Biomonitoring, Artificial Intelligence, and Precision Space Health}, author = {{Ryan T. Scott} and {E. Antonsen} and {L. Sanders} and {Jaden J. A. Hastings} and {Seung-min Park} and {Graham Mackintosh} and {R. Reynolds} and {A. Hoarfrost} and {A. Sawyer} and {C. Greene} and {Benjamin S. Glicksberg} and {C. Theriot} and {D. Berrios} and {Jack M. Miller} and {Joel Babdor} and {Richard Barker} and {S. Baranzini} and {Afshin Beheshti} and {S. Chalk} and {Guillermo M. Delgado-Aparicio} and {M. Haendel} and {Arif A. Hamid} and {P. Heller} and {Daniel Jamieson} and {K. Jarvis} and {John Kalantari} and {K. Khezeli} and {S. Komarova} and {M. Komorowski} and {Prachi Kothiyal} and {A. Mahabal} and {U. Manor} and {H. Martín} and {Christopher E. Mason} and {Mona Matar} and {G. Mias} and {J. Myers} and {Jr} and {Charlotte A. Nelson} and {Jonathan Oribello} and {P. Parsons-Wingerter} and {R. Prabhu} and {A. Qutub} and {J. Rask} and {Amanda M. Saravia-Butler} and {S. Saria} and {N. Singh} and {Frank Soboczenski} and {M. Snyder} and {Karthik Soman} and {D. V. Valen} and {K. Venkateswaran} and {L. Warren} and {Liz Worthey} and {Jason H. Yang} and {M. Zitnik} and {S. V. C. Kbr} and {Space Biosciences Division} and {N. R. Center} and {M. Field} and {Ca} and {USA.} and {Department of Preventive Medicine} and {Center for Individualized Medicine} and {Baylor College of Medicine} and {Houston} and {Tx} and {Blue Marble Space Institute of Science} and {D. Physiology} and {Biophysics} and {Weill Cornell Medicine} and {N. York} and {Ny} and {D. Urology} and {D. Radiology} and {S. Medicine} and {Stanford} and {Bay Area Environmental Research Institute} and {Mortality ResearchConsulting} and {Inc.} and {Universities Space Research Association} and {UC Space Health} and {D. Surgery} and {U. California} and {San Francisco} and {AI CenterforHealth} and {D. Biochemistry} and {Molecular Genetics} and {University of Maryland School of Medicine} and {Anschutz Medical Campus} and {Aurora} and {Co} and {Hasso Plattner Institute for Digital Health at Mount Sinai} and {Department of Genetics} and {Genomic Sciences} and {I. A. Sinai} and {Department of Preventive Medicine} and {C. Health} and {Utmb} and {Galveston} and {Human Health} and {Performance Directorate} and {NASAMarshall Space Flight Center} and {D. Microbiology} and {Immunology} and {Department of Otolaryngology} and {Head} and {N. Surgery} and {University of San Francisco} and {The Gilroy AstroBiology Research Group} and {The University of Wisconsin} and {Madison} and {Wi} and {Weill Institute for Neurosciences} and {D. Neurology} and {D. Chemistry} and {U. Florida} and {Jacksonville} and {Fl} and {D. Analytics} and {G. I. O. Technology} and {Lima} and {Perú} and {Department of Neuroscience} and {U. Minnesota} and {Minneapolis} and {Mn} and {Department of Materials Science} and {College of Materials Science} and {San Diego State University} and {San José} and {Biorelate} and {Manchester} and {United Kingdom.} and {Center for Individualized Medicine} and {D. Surgery} and {D. Sciences} and {Mayo Clinic} and {Rochester} and {Faculty of Veterinary Medicine} and {Oral Health Sciences} and {M. University} and {Montreal.} and {Quebec.} and {Canada.} and {Faculty of Veterinary Medicine} and {Cancer} and {I. -. London} and {London} and {SymbioSeq Llc} and {Ashburn} and {Va} and {Center for Data Driven Discovery} and {C. I. O. Technology.} and {Pasadena} and {Waitt Advanced Biophotonics Center} and {Chan-Zuckerberg Imaging Scientist Fellow} and {Salk Institute for Biological Studies} and {La Jolla} and {Biological Systems} and {Engineering Division} and {Lawrence Berkeley National Lab.} and {Berkeley} and {Doe Agile BioFoundry} and {Emeryville} and {Joint BioEnergy Institute} and {Human Research Program Cross-cutting Computational Model Project} and {N. R. Center} and {Cleveland} and {Oh} and {Institute for Computational Science} and {Engineering} and {M. Biology} and {M. University} and {E. Lansing.} and {Mi} and {Low Exploration Gravity Technology} and {AI Matrix Consortium} and {Department of Biomedical Engineering} and {U. Texas} and {S. Antonio} and {UT Health Sciences} and {Office of the Director} and {Logyx} and {Computer Science} and {Statistics} and {H. Policy} and {J. University} and {Baltimore.} and {Md.} and {Ml} and {Ai} and {H. Lab} and {B. Health} and {Biotechnology} and {Planetary Protection Group} and {J. P. Laboratory} and {Sphes} and {Medical Faculty} and {King’s College London} and {S. Medicine} and {Department of Biology} and {Iss National Laboratory} and {Center for Space} and {Melbourne} and {Uab Center for Computational Biology} and {Data Science} and {U. Alabama} and {Birmingham} and {Al} and {Center for Emerging} and {Re-Emerging Pathogens} and {Biochemistry} and {Rutgers New Jersey Medical School} and {Newark} and {Nj} and {Department of Biomedical Informatics} and {Harvard Medical School} and {Harvard Data Science} and {Broad Institute of Mit} and {Harvard} and {Harvard University} and {Boston} and {Ma.}}, year = 2021, month = {12}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/0d6d142dc49cf7537ece045d8d469fd014a5d3b6}, } - Zach Wood-Doughty, Isabel Cachola, and Mark Dredze, “Proxy Model Explanations for Time Series RNNs,” in International Conference on Machine Learning and Applications, 2021.
[BibTeX] [Link]@inproceedings{246291268, title = {Proxy Model Explanations for Time Series RNNs}, author = {{Zach Wood-Doughty} and {Isabel Cachola} and {Mark Dredze}}, year = 2021, month = {12}, booktitle = {International Conference on Machine Learning and Applications}, url = {https://www.semanticscholar.org/paper/9e031c15797f9e41598a6c7ebe583e3bb72dceb0}, } - R. Shin, C. Lin, S. Thomson, C. Chen, S. Roy, E. A. Platanios, A. Pauls, D. Klein, J. Eisner, and B. Van Durme, “Constrained Language Models Yield Few-Shot Semantic Parsers,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 7699–7715. doi:10.18653/v1/2021.emnlp-main.608
[BibTeX] [Abstract] [Link]
We explore the use of large pretrained language models as few-shot semantic parsers. The goal in semantic parsing is to generate a structured meaning representation given a natural language input. However, language models are trained to generate natural language. To bridge the gap, we use language models to paraphrase inputs into a controlled sublanguage resembling English that can be automatically mapped to a target meaning representation. Our results demonstrate that with only a small amount of data and very little code to convert into English-like representations, our blueprint for rapidly bootstrapping semantic parsers leads to surprisingly effective performance on multiple community tasks, greatly exceeding baseline methods also trained on the same limited data.
@inproceedings{shin-etal-2021-constrained, title = "Constrained Language Models Yield Few-Shot Semantic Parsers", author = "Shin, Richard and Lin, Christopher and Thomson, Sam and Chen, Charles and Roy, Subhro and Platanios, Emmanouil Antonios and Pauls, Adam and Klein, Dan and Eisner, Jason and Van Durme, Benjamin", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.608", doi = "10.18653/v1/2021.emnlp-main.608", pages = "7699--7715", abstract = "We explore the use of large pretrained language models as few-shot semantic parsers. The goal in semantic parsing is to generate a structured meaning representation given a natural language input. However, language models are trained to generate natural language. To bridge the gap, we use language models to paraphrase inputs into a controlled sublanguage resembling English that can be automatically mapped to a target meaning representation. Our results demonstrate that with only a small amount of data and very little code to convert into English-like representations, our blueprint for rapidly bootstrapping semantic parsers leads to surprisingly effective performance on multiple community tasks, greatly exceeding baseline methods also trained on the same limited data.", } - F. Akhbardeh, A. Arkhangorodsky, M. Biesialska, O. Bojar, R. Chatterjee, V. Chaudhary, M. R. Costa-jussa, C. España-Bonet, A. Fan, C. Federmann, M. Freitag, Y. Graham, R. Grundkiewicz, B. Haddow, L. Harter, K. Heafield, C. Homan, M. Huck, K. Amponsah-Kaakyire, J. Kasai, D. Khashabi, K. Knight, T. Kocmi, P. Koehn, N. Lourie, C. Monz, M. Morishita, M. Nagata, A. Nagesh, T. Nakazawa, M. Negri, S. Pal, A. A. Tapo, M. Turchi, V. Vydrin, and M. Zampieri, “Findings of the 2021 Conference on Machine Translation (WMT21),” in Proceedings of the Sixth Conference on Machine Translation, Online, 2021, p. 1–88.
[BibTeX] [Abstract] [Link]
This paper presents the results of the newstranslation task, the multilingual low-resourcetranslation for Indo-European languages, thetriangular translation task, and the automaticpost-editing task organised as part of the Con-ference on Machine Translation (WMT) 2021.In the news task, participants were asked tobuild machine translation systems for any of10 language pairs, to be evaluated on test setsconsisting mainly of news stories. The taskwas also opened up to additional test suites toprobe specific aspects of translation.
@inproceedings{akhbardeh-etal-2021-findings, title = "Findings of the 2021 Conference on Machine Translation ({WMT}21)", author = "Akhbardeh, Farhad and Arkhangorodsky, Arkady and Biesialska, Magdalena and Bojar, Ond{\v{r}}ej and Chatterjee, Rajen and Chaudhary, Vishrav and Costa-jussa, Marta R. and Espa{\~n}a-Bonet, Cristina and Fan, Angela and Federmann, Christian and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Haddow, Barry and Harter, Leonie and Heafield, Kenneth and Homan, Christopher and Huck, Matthias and Amponsah-Kaakyire, Kwabena and Kasai, Jungo and Khashabi, Daniel and Knight, Kevin and Kocmi, Tom and Koehn, Philipp and Lourie, Nicholas and Monz, Christof and Morishita, Makoto and Nagata, Masaaki and Nagesh, Ajay and Nakazawa, Toshiaki and Negri, Matteo and Pal, Santanu and Tapo, Allahsera Auguste and Turchi, Marco and Vydrin, Valentin and Zampieri, Marcos", editor = "Barrault, Loic and Bojar, Ondrej and Bougares, Fethi and Chatterjee, Rajen and Costa-jussa, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Yepes, Antonio Jimeno and Koehn, Philipp and Kocmi, Tom and Martins, Andre and Morishita, Makoto and Monz, Christof", booktitle = "Proceedings of the Sixth Conference on Machine Translation", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wmt-1.1", pages = "1--88", abstract = "This paper presents the results of the newstranslation task, the multilingual low-resourcetranslation for Indo-European languages, thetriangular translation task, and the automaticpost-editing task organised as part of the Con-ference on Machine Translation (WMT) 2021.In the news task, participants were asked tobuild machine translation systems for any of10 language pairs, to be evaluated on test setsconsisting mainly of news stories. The taskwas also opened up to additional test suites toprobe specific aspects of translation.", } - Junfei Xiao, Longlong Jing, Lin Zhang, Ju He, Qi She, Zongwei Zhou, A. Yuille, and Yingwei Li, “Learning from Temporal Gradient for Semi-supervised Action Recognition,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{244709803, title = {Learning from Temporal Gradient for Semi-supervised Action Recognition}, author = {{Junfei Xiao} and {Longlong Jing} and {Lin Zhang} and {Ju He} and {Qi She} and {Zongwei Zhou} and {A. Yuille} and {Yingwei Li}}, year = 2021, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/069e9bb3c9674441c6872767f33ae5d9a4931cd3}, } - Pengfei Guo and Vishal M. Patel, “Reference-based Magnetic Resonance Image Reconstruction Using Texture Transforme,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{244345634, title = {Reference-based Magnetic Resonance Image Reconstruction Using Texture Transforme}, author = {{Pengfei Guo} and {Vishal M. Patel}}, year = 2021, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7bab95180b52749d2b018d120d8f04bba520ee0f}, } - Jinghao Zhou, Chen Wei, Huiyu Wang, Wei Shen, Cihang Xie, A. Yuille, and Tao Kong, “iBOT: Image BERT Pre-Training with Online Tokenizer,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{244117494, title = {iBOT: Image BERT Pre-Training with Online Tokenizer}, author = {{Jinghao Zhou} and {Chen Wei} and {Huiyu Wang} and {Wei Shen} and {Cihang Xie} and {A. Yuille} and {Tao Kong}}, year = 2021, month = {11}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/9653c070724e44f023e8cc3ec79f0b9e6d59480d}, } - Yutong Bai, Jieru Mei, A. Yuille, and Cihang Xie, “Are Transformers More Robust Than CNNs?,” in Neural Information Processing Systems, 2021.
[BibTeX] [Link]@inproceedings{243938451, title = {Are Transformers More Robust Than CNNs?}, author = {{Yutong Bai} and {Jieru Mei} and {A. Yuille} and {Cihang Xie}}, year = 2021, month = {11}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/35c0800e657faa18cf3fc3629bdbeafbb976b006}, } - Yu Zeng, Zhe L. Lin, and Vishal M. Patel, “SketchEdit: Mask-Free Local Image Manipulation with Partial Sketches,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{244729626, title = {SketchEdit: Mask-Free Local Image Manipulation with Partial Sketches}, author = {{Yu Zeng} and {Zhe L. Lin} and {Vishal M. Patel}}, year = 2021, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/378aa9ad054989663c6db5f2fe90d6982340e28b}, } - Bingchen Zhao, Shaozuo Yu, Wufei Ma, M. Yu, Shenxiao Mei, Angtian Wang, Ju He, A. Yuille, and Adam Kortylewski, “OOD-CV: A Benchmark for Robustness to Out-of-Distribution Shifts of Individual Nuisances in Natural Images,” in European Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{251041144, title = {OOD-CV: A Benchmark for Robustness to Out-of-Distribution Shifts of Individual Nuisances in Natural Images}, author = {{Bingchen Zhao} and {Shaozuo Yu} and {Wufei Ma} and {M. Yu} and {Shenxiao Mei} and {Angtian Wang} and {Ju He} and {A. Yuille} and {Adam Kortylewski}}, year = 2021, month = {11}, booktitle = {European Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/8f693bc2219607316e143ba543ae0e7abca6a4b1}, } - M. Yarmohammadi, S. Wu, M. Marone, H. Xu, S. Ebner, G. Qin, Y. Chen, J. Guo, C. Harman, K. Murray, A. S. White, M. Dredze, and B. Van Durme, “Everything Is All It Takes: A Multipronged Strategy for Zero-Shot Cross-Lingual Information Extraction,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 1950–1967. doi:10.18653/v1/2021.emnlp-main.149
[BibTeX] [Abstract] [Link]
Zero-shot cross-lingual information extraction (IE) describes the construction of an IE model for some target language, given existing annotations exclusively in some other language, typically English. While the advance of pretrained multilingual encoders suggests an easy optimism of {“}train on English, run on any language{”}, we find through a thorough exploration and extension of techniques that a combination of approaches, both new and old, leads to better performance than any one cross-lingual strategy in particular. We explore techniques including data projection and self-training, and how different pretrained encoders impact them. We use English-to-Arabic IE as our initial example, demonstrating strong performance in this setting for event extraction, named entity recognition, part-of-speech tagging, and dependency parsing. We then apply data projection and self-training to three tasks across eight target languages. Because no single set of techniques performs the best across all tasks, we encourage practitioners to explore various configurations of the techniques described in this work when seeking to improve on zero-shot training.
@inproceedings{yarmohammadi-etal-2021-everything, title = "Everything Is All It Takes: A Multipronged Strategy for Zero-Shot Cross-Lingual Information Extraction", author = "Yarmohammadi, Mahsa and Wu, Shijie and Marone, Marc and Xu, Haoran and Ebner, Seth and Qin, Guanghui and Chen, Yunmo and Guo, Jialiang and Harman, Craig and Murray, Kenton and White, Aaron Steven and Dredze, Mark and Van Durme, Benjamin", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.149", doi = "10.18653/v1/2021.emnlp-main.149", pages = "1950--1967", abstract = "Zero-shot cross-lingual information extraction (IE) describes the construction of an IE model for some target language, given existing annotations exclusively in some other language, typically English. While the advance of pretrained multilingual encoders suggests an easy optimism of {``}train on English, run on any language{''}, we find through a thorough exploration and extension of techniques that a combination of approaches, both new and old, leads to better performance than any one cross-lingual strategy in particular. We explore techniques including data projection and self-training, and how different pretrained encoders impact them. We use English-to-Arabic IE as our initial example, demonstrating strong performance in this setting for event extraction, named entity recognition, part-of-speech tagging, and dependency parsing. We then apply data projection and self-training to three tasks across eight target languages. Because no single set of techniques performs the best across all tasks, we encourage practitioners to explore various configurations of the techniques described in this work when seeking to improve on zero-shot training.", } - E. Salesky, D. Etter, and M. Post, “Robust Open-Vocabulary Translation from Visual Text Representations,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 7235–7252. doi:10.18653/v1/2021.emnlp-main.576
[BibTeX] [Abstract] [Link]
Machine translation models have discrete vocabularies and commonly use subword segmentation techniques to achieve an {`}open vocabulary.{‘} This approach relies on consistent and correct underlying unicode sequences, and makes models susceptible to degradation from common types of noise and variation. Motivated by the robustness of human language processing, we propose the use of visual text representations, which dispense with a finite set of text embeddings in favor of continuous vocabularies created by processing visually rendered text with sliding windows. We show that models using visual text representations approach or match performance of traditional text models on small and larger datasets. More importantly, models with visual embeddings demonstrate significant robustness to varied types of noise, achieving e.g., 25.9 BLEU on a character permuted German{–}English task where subword models degrade to 1.9.
@inproceedings{salesky-etal-2021-robust, title = "Robust Open-Vocabulary Translation from Visual Text Representations", author = "Salesky, Elizabeth and Etter, David and Post, Matt", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.576", doi = "10.18653/v1/2021.emnlp-main.576", pages = "7235--7252", abstract = "Machine translation models have discrete vocabularies and commonly use subword segmentation techniques to achieve an {`}open vocabulary.{'} This approach relies on consistent and correct underlying unicode sequences, and makes models susceptible to degradation from common types of noise and variation. Motivated by the robustness of human language processing, we propose the use of visual text representations, which dispense with a finite set of text embeddings in favor of continuous vocabularies created by processing visually rendered text with sliding windows. We show that models using visual text representations approach or match performance of traditional text models on small and larger datasets. More importantly, models with visual embeddings demonstrate significant robustness to varied types of noise, achieving e.g., 25.9 BLEU on a character permuted German{--}English task where subword models degrade to 1.9.", } - Jeya Maria Jose Valanarasu, R. Yasarla, and Vishal M. Patel, “TransWeather: Transformer-based Restoration of Images Degraded by Adverse Weather Conditions,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{244714491, title = {TransWeather: Transformer-based Restoration of Images Degraded by Adverse Weather Conditions}, author = {{Jeya Maria Jose Valanarasu} and {R. Yasarla} and {Vishal M. Patel}}, year = 2021, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/b27d3be4264dcd06f990b44968f4382526f24f1e}, } - M. M. I. Alam, I. Kvapil{‘i}ková, A. Anastasopoulos, L. Besacier, G. Dinu, M. Federico, M. Gallé, K. Jung, P. Koehn, and V. Nikoulina, “Findings of the WMT Shared Task on Machine Translation Using Terminologies,” in Proceedings of the Sixth Conference on Machine Translation, Online, 2021, p. 652–663.
[BibTeX] [Abstract] [Link]
Language domains that require very careful use of terminology are abundant and reflect a significant part of the translation industry. In this work we introduce a benchmark for evaluating the quality and consistency of terminology translation, focusing on the medical (and COVID-19 specifically) domain for five language pairs: English to French, Chinese, Russian, and Korean, as well as Czech to German. We report the descriptions and results of the participating systems, commenting on the need for further research efforts towards both more adequate handling of terminologies as well as towards a proper formulation and evaluation of the task.
@inproceedings{alam-etal-2021-findings, title = "Findings of the {WMT} Shared Task on Machine Translation Using Terminologies", author = "Alam, Md Mahfuz Ibn and Kvapil{\'\i}kov{\'a}, Ivana and Anastasopoulos, Antonios and Besacier, Laurent and Dinu, Georgiana and Federico, Marcello and Gall{\'e}, Matthias and Jung, Kweonwoo and Koehn, Philipp and Nikoulina, Vassilina", editor = "Barrault, Loic and Bojar, Ondrej and Bougares, Fethi and Chatterjee, Rajen and Costa-jussa, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Yepes, Antonio Jimeno and Koehn, Philipp and Kocmi, Tom and Martins, Andre and Morishita, Makoto and Monz, Christof", booktitle = "Proceedings of the Sixth Conference on Machine Translation", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wmt-1.69", pages = "652--663", abstract = "Language domains that require very careful use of terminology are abundant and reflect a significant part of the translation industry. In this work we introduce a benchmark for evaluating the quality and consistency of terminology translation, focusing on the medical (and COVID-19 specifically) domain for five language pairs: English to French, Chinese, Russian, and Korean, as well as Czech to German. We report the descriptions and results of the participating systems, commenting on the need for further research efforts towards both more adequate handling of terminologies as well as towards a proper formulation and evaluation of the task.", } - Huaijin Pi, Huiyu Wang, Yingwei Li, Zizhang Li, and A. Yuille, “Searching for TrioNet: Combining Convolution with Local and Global Self-Attention,” in British Machine Vision Conference, 2021.
[BibTeX] [Link]@inproceedings{244117374, title = {Searching for TrioNet: Combining Convolution with Local and Global Self-Attention}, author = {{Huaijin Pi} and {Huiyu Wang} and {Yingwei Li} and {Zizhang Li} and {A. Yuille}}, year = 2021, month = {11}, booktitle = {British Machine Vision Conference}, url = {https://www.semanticscholar.org/paper/2ecdb624c2a87624e27c34e3af388b559a0ba06c}, } - Jiyang Qi, Yan Gao, Yao Hu, Xinggang Wang, Xiaoyu Liu, Xiang Bai, S. Belongie, A. Yuille, Philip H. S. Torr, and S. Bai, “Occluded Video Instance Segmentation: Dataset and ICCV 2021 Challenge,” in NeurIPS Datasets and Benchmarks, 2021.
[BibTeX] [Link]@inproceedings{244117621, title = {Occluded Video Instance Segmentation: Dataset and ICCV 2021 Challenge}, author = {{Jiyang Qi} and {Yan Gao} and {Yao Hu} and {Xinggang Wang} and {Xiaoyu Liu} and {Xiang Bai} and {S. Belongie} and {A. Yuille} and {Philip H. S. Torr} and {S. Bai}}, year = 2021, month = {11}, booktitle = {NeurIPS Datasets and Benchmarks}, url = {https://www.semanticscholar.org/paper/60b137e3b5f378e50d7875bb5ad0390d107374bb}, } - A. Buczak, Benjamin D. Baugher, Christine S. Martin, Meg W. Keiley-Listermann, J. Howard, Nathan H. Parrish, Anton Q. Stalick, D. S. Berman, and Mark Dredze, “Crystal Cube: Forecasting Disruptive Events,” in Applied Artificial Intelligence, 2021.
[BibTeX] [Link]@inproceedings{244096848, title = {Crystal Cube: Forecasting Disruptive Events}, author = {{A. Buczak} and {Benjamin D. Baugher} and {Christine S. Martin} and {Meg W. Keiley-Listermann} and {J. Howard} and {Nathan H. Parrish} and {Anton Q. Stalick} and {D. S. Berman} and {Mark Dredze}}, year = 2021, month = {11}, booktitle = {Applied Artificial Intelligence}, url = {https://www.semanticscholar.org/paper/3168dec5c6a5c1441f258c14d05f8520f20ecbaf}, } - S. Ding, M. Junczys-Dowmunt, M. Post, and P. Koehn, “Levenshtein Training for Word-level Quality Estimation,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 6724–6733. doi:10.18653/v1/2021.emnlp-main.539
[BibTeX] [Abstract] [Link]
We propose a novel scheme to use the Levenshtein Transformer to perform the task of word-level quality estimation. A Levenshtein Transformer is a natural fit for this task: trained to perform decoding in an iterative manner, a Levenshtein Transformer can learn to post-edit without explicit supervision. To further minimize the mismatch between the translation task and the word-level QE task, we propose a two-stage transfer learning procedure on both augmented data and human post-editing data. We also propose heuristics to construct reference labels that are compatible with subword-level finetuning and inference. Results on WMT 2020 QE shared task dataset show that our proposed method has superior data efficiency under the data-constrained setting and competitive performance under the unconstrained setting.
@inproceedings{ding-etal-2021-levenshtein, title = "{L}evenshtein Training for Word-level Quality Estimation", author = "Ding, Shuoyang and Junczys-Dowmunt, Marcin and Post, Matt and Koehn, Philipp", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.539", doi = "10.18653/v1/2021.emnlp-main.539", pages = "6724--6733", abstract = "We propose a novel scheme to use the Levenshtein Transformer to perform the task of word-level quality estimation. A Levenshtein Transformer is a natural fit for this task: trained to perform decoding in an iterative manner, a Levenshtein Transformer can learn to post-edit without explicit supervision. To further minimize the mismatch between the translation task and the word-level QE task, we propose a two-stage transfer learning procedure on both augmented data and human post-editing data. We also propose heuristics to construct reference labels that are compatible with subword-level finetuning and inference. Results on WMT 2020 QE shared task dataset show that our proposed method has superior data efficiency under the data-constrained setting and competitive performance under the unconstrained setting.", } - A. Chinta, J. Zhang, A. DeLucia, M. Dredze, and A. L. Buczak, “Study of Manifestation of Civil Unrest on Twitter,” in Proceedings of the Seventh Workshop on Noisy User-generated Text (W-NUT 2021), Online, 2021, p. 396–409. doi:10.18653/v1/2021.wnut-1.44
[BibTeX] [Abstract] [Link]
Twitter is commonly used for civil unrest detection and forecasting tasks, but there is a lack of work in evaluating \textit{how} civil unrest manifests on Twitter across countries and events. We present two in-depth case studies for two specific large-scale events, one in a country with high (English) Twitter usage (Johannesburg riots in South Africa) and one in a country with low Twitter usage (Burayu massacre protests in Ethiopia). We show that while there is event signal during the events, there is little signal leading up to the events. In addition to the case studies, we train Ngram-based models on a larger set of Twitter civil unrest data across time, events, and countries and use machine learning explainability tools (SHAP) to identify important features. The models were able to find words indicative of civil unrest that generalized across countries. The 42 countries span Africa, Middle East, and Southeast Asia and the events range occur between 2014 and 2019.
@inproceedings{chinta-etal-2021-study, title = "Study of Manifestation of Civil Unrest on {T}witter", author = "Chinta, Abhinav and Zhang, Jingyu and DeLucia, Alexandra and Dredze, Mark and Buczak, Anna L.", editor = "Xu, Wei and Ritter, Alan and Baldwin, Tim and Rahimi, Afshin", booktitle = "Proceedings of the Seventh Workshop on Noisy User-generated Text (W-NUT 2021)", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wnut-1.44", doi = "10.18653/v1/2021.wnut-1.44", pages = "396--409", abstract = "Twitter is commonly used for civil unrest detection and forecasting tasks, but there is a lack of work in evaluating \textit{how} civil unrest manifests on Twitter across countries and events. We present two in-depth case studies for two specific large-scale events, one in a country with high (English) Twitter usage (Johannesburg riots in South Africa) and one in a country with low Twitter usage (Burayu massacre protests in Ethiopia). We show that while there is event signal during the events, there is little signal leading up to the events. In addition to the case studies, we train Ngram-based models on a larger set of Twitter civil unrest data across time, events, and countries and use machine learning explainability tools (SHAP) to identify important features. The models were able to find words indicative of civil unrest that generalized across countries. The 42 countries span Africa, Middle East, and Southeast Asia and the events range occur between 2014 and 2019.", } - G. Kumar, P. Koehn, and S. Khudanpur, “Learning Feature Weights using Reward Modeling for Denoising Parallel Corpora,” in Proceedings of the Sixth Conference on Machine Translation, Online, 2021, p. 1100–1109.
[BibTeX] [Abstract] [Link]
Large web-crawled corpora represent an excellent resource for improving the performance of Neural Machine Translation (NMT) systems across several language pairs. However, since these corpora are typically extremely noisy, their use is fairly limited. Current approaches to deal with this problem mainly focus on filtering using heuristics or single features such as language model scores or bi-lingual similarity. This work presents an alternative approach which learns weights for multiple sentence-level features. These feature weights which are optimized directly for the task of improving translation performance, are used to score and filter sentences in the noisy corpora more effectively. We provide results of applying this technique to building NMT systems using the Paracrawl corpus for Estonian-English and show that it beats strong single feature baselines and hand designed combinations. Additionally, we analyze the sensitivity of this method to different types of noise and explore if the learned weights generalize to other language pairs using the Maltese-English Paracrawl corpus.
@inproceedings{kumar-etal-2021-learning-feature, title = "Learning Feature Weights using Reward Modeling for Denoising Parallel Corpora", author = "Kumar, Gaurav and Koehn, Philipp and Khudanpur, Sanjeev", editor = "Barrault, Loic and Bojar, Ondrej and Bougares, Fethi and Chatterjee, Rajen and Costa-jussa, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Yepes, Antonio Jimeno and Koehn, Philipp and Kocmi, Tom and Martins, Andre and Morishita, Makoto and Monz, Christof", booktitle = "Proceedings of the Sixth Conference on Machine Translation", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wmt-1.118", pages = "1100--1109", abstract = "Large web-crawled corpora represent an excellent resource for improving the performance of Neural Machine Translation (NMT) systems across several language pairs. However, since these corpora are typically extremely noisy, their use is fairly limited. Current approaches to deal with this problem mainly focus on filtering using heuristics or single features such as language model scores or bi-lingual similarity. This work presents an alternative approach which learns weights for multiple sentence-level features. These feature weights which are optimized directly for the task of improving translation performance, are used to score and filter sentences in the noisy corpora more effectively. We provide results of applying this technique to building NMT systems using the Paracrawl corpus for Estonian-English and show that it beats strong single feature baselines and hand designed combinations. Additionally, we analyze the sensitivity of this method to different types of noise and explore if the learned weights generalize to other language pairs using the Maltese-English Paracrawl corpus.", } - H. Xu, B. Van Durme, and K. Murray, “BERT, mBERT, or BiBERT? A Study on Contextualized Embeddings for Neural Machine Translation,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 6663–6675. doi:10.18653/v1/2021.emnlp-main.534
[BibTeX] [Abstract] [Link]
The success of bidirectional encoders using masked language models, such as BERT, on numerous natural language processing tasks has prompted researchers to attempt to incorporate these pre-trained models into neural machine translation (NMT) systems. However, proposed methods for incorporating pre-trained models are non-trivial and mainly focus on BERT, which lacks a comparison of the impact that other pre-trained models may have on translation performance. In this paper, we demonstrate that simply using the output (contextualized embeddings) of a tailored and suitable bilingual pre-trained language model (dubbed BiBERT) as the input of the NMT encoder achieves state-of-the-art translation performance. Moreover, we also propose a stochastic layer selection approach and a concept of a dual-directional translation model to ensure the sufficient utilization of contextualized embeddings. In the case of without using back translation, our best models achieve BLEU scores of 30.45 for EnâDe and 38.61 for DeâEn on the IWSLT{‘}14 dataset, and 31.26 for EnâDe and 34.94 for DeâEn on the WMT{‘}14 dataset, which exceeds all published numbers.
@inproceedings{xu-etal-2021-bert, title = "{BERT}, m{BERT}, or {B}i{BERT}? A Study on Contextualized Embeddings for Neural Machine Translation", author = "Xu, Haoran and Van Durme, Benjamin and Murray, Kenton", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.534", doi = "10.18653/v1/2021.emnlp-main.534", pages = "6663--6675", abstract = "The success of bidirectional encoders using masked language models, such as BERT, on numerous natural language processing tasks has prompted researchers to attempt to incorporate these pre-trained models into neural machine translation (NMT) systems. However, proposed methods for incorporating pre-trained models are non-trivial and mainly focus on BERT, which lacks a comparison of the impact that other pre-trained models may have on translation performance. In this paper, we demonstrate that simply using the output (contextualized embeddings) of a tailored and suitable bilingual pre-trained language model (dubbed BiBERT) as the input of the NMT encoder achieves state-of-the-art translation performance. Moreover, we also propose a stochastic layer selection approach and a concept of a dual-directional translation model to ensure the sufficient utilization of contextualized embeddings. In the case of without using back translation, our best models achieve BLEU scores of 30.45 for EnâDe and 38.61 for DeâEn on the IWSLT{'}14 dataset, and 31.26 for EnâDe and 34.94 for DeâEn on the WMT{'}14 dataset, which exceeds all published numbers.", } - Tiange Xiang, Yixiao Zhang, Yongyi Lu, A. Yuille, Chaoyi Zhang, Weidong (Tom) Cai, and Zongwei Zhou, “SQUID: Deep Feature In-Painting for Unsupervised Anomaly Detection,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{257766829, title = {SQUID: Deep Feature In-Painting for Unsupervised Anomaly Detection}, author = {{Tiange Xiang} and {Yixiao Zhang} and {Yongyi Lu} and {A. Yuille} and {Chaoyi Zhang} and {Weidong (Tom) Cai} and {Zongwei Zhou}}, year = 2021, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/e2977c67f55b8a2a58ff1c232c96bed25002f8a2}, } - S. Ding, M. Junczys-Dowmunt, M. Post, C. Federmann, and P. Koehn, “The JHU-Microsoft Submission for WMT21 Quality Estimation Shared Task,” in Proceedings of the Sixth Conference on Machine Translation, Online, 2021, p. 904–910.
[BibTeX] [Abstract] [Link]
This paper presents the JHU-Microsoft joint submission for WMT 2021 quality estimation shared task. We only participate in Task 2 (post-editing effort estimation) of the shared task, focusing on the target-side word-level quality estimation. The techniques we experimented with include Levenshtein Transformer training and data augmentation with a combination of forward, backward, round-trip translation, and pseudo post-editing of the MT output. We demonstrate the competitiveness of our system compared to the widely adopted OpenKiwi-XLM baseline. Our system is also the top-ranking system on the MT MCC metric for the English-German language pair.
@inproceedings{ding-etal-2021-jhu, title = "The {JHU}-{M}icrosoft Submission for {WMT}21 Quality Estimation Shared Task", author = "Ding, Shuoyang and Junczys-Dowmunt, Marcin and Post, Matt and Federmann, Christian and Koehn, Philipp", editor = "Barrault, Loic and Bojar, Ondrej and Bougares, Fethi and Chatterjee, Rajen and Costa-jussa, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Yepes, Antonio Jimeno and Koehn, Philipp and Kocmi, Tom and Martins, Andre and Morishita, Makoto and Monz, Christof", booktitle = "Proceedings of the Sixth Conference on Machine Translation", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wmt-1.94", pages = "904--910", abstract = "This paper presents the JHU-Microsoft joint submission for WMT 2021 quality estimation shared task. We only participate in Task 2 (post-editing effort estimation) of the shared task, focusing on the target-side word-level quality estimation. The techniques we experimented with include Levenshtein Transformer training and data augmentation with a combination of forward, backward, round-trip translation, and pseudo post-editing of the MT output. We demonstrate the competitiveness of our system compared to the widely adopted OpenKiwi-XLM baseline. Our system is also the top-ranking system on the MT MCC metric for the English-German language pair.", } - M. A. Gordon, K. Duh, and J. Kaplan, “Data and Parameter Scaling Laws for Neural Machine Translation,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 5915–5922. doi:10.18653/v1/2021.emnlp-main.478
[BibTeX] [Abstract] [Link]
We observe that the development cross-entropy loss of supervised neural machine translation models scales like a power law with the amount of training data and the number of non-embedding parameters in the model. We discuss some practical implications of these results, such as predicting BLEU achieved by large scale models and predicting the ROI of labeling data in low-resource language pairs.
@inproceedings{gordon-etal-2021-data, title = "Data and Parameter Scaling Laws for Neural Machine Translation", author = "Gordon, Mitchell A and Duh, Kevin and Kaplan, Jared", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.478", doi = "10.18653/v1/2021.emnlp-main.478", pages = "5915--5922", abstract = "We observe that the development cross-entropy loss of supervised neural machine translation models scales like a power law with the amount of training data and the number of non-embedding parameters in the model. We discuss some practical implications of these results, such as predicting BLEU achieved by large scale models and predicting the ROI of labeling data in low-resource language pairs.", } - P. Xia and B. Van Durme, “Moving on from OntoNotes: Coreference Resolution Model Transfer,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Online and Punta Cana, Dominican Republic, 2021, p. 5241–5256. doi:10.18653/v1/2021.emnlp-main.425
[BibTeX] [Abstract] [Link]
Academic neural models for coreference resolution (coref) are typically trained on a single dataset, OntoNotes, and model improvements are benchmarked on that same dataset. However, real-world applications of coref depend on the annotation guidelines and the domain of the target dataset, which often differ from those of OntoNotes. We aim to quantify transferability of coref models based on the number of annotated documents available in the target dataset. We examine eleven target datasets and find that continued training is consistently effective and especially beneficial when there are few target documents. We establish new benchmarks across several datasets, including state-of-the-art results on PreCo.
@inproceedings{xia-van-durme-2021-moving, title = "Moving on from {O}nto{N}otes: Coreference Resolution Model Transfer", author = "Xia, Patrick and Van Durme, Benjamin", editor = "Moens, Marie-Francine and Huang, Xuanjing and Specia, Lucia and Yih, Scott Wen-tau", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", month = nov, year = "2021", address = "Online and Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.emnlp-main.425", doi = "10.18653/v1/2021.emnlp-main.425", pages = "5241--5256", abstract = "Academic neural models for coreference resolution (coref) are typically trained on a single dataset, OntoNotes, and model improvements are benchmarked on that same dataset. However, real-world applications of coref depend on the annotation guidelines and the domain of the target dataset, which often differ from those of OntoNotes. We aim to quantify transferability of coref models based on the number of annotated documents available in the target dataset. We examine eleven target datasets and find that continued training is consistently effective and especially beneficial when there are few target documents. We establish new benchmarks across several datasets, including state-of-the-art results on PreCo.", } - C. Tran, S. Bhosale, J. Cross, P. Koehn, S. Edunov, and A. Fan, “Facebook AI’s WMT21 News Translation Task Submission,” in Proceedings of the Sixth Conference on Machine Translation, Online, 2021, p. 205–215.
[BibTeX] [Abstract] [Link]
We describe Facebook{‘}s multilingual model submission to the WMT2021 shared task on news translation. We participate in 14 language directions: English to and from Czech, German, Hausa, Icelandic, Japanese, Russian, and Chinese. To develop systems covering all these directions, we focus on multilingual models. We utilize data from all available sources {–-} WMT, large-scale data mining, and in-domain backtranslation {–-} to create high quality bilingual and multilingual baselines. Subsequently, we investigate strategies for scaling multilingual model size, such that one system has sufficient capacity for high quality representations of all eight languages. Our final submission is an ensemble of dense and sparse Mixture-of-Expert multilingual translation models, followed by finetuning on in-domain news data and noisy channel reranking. Compared to previous year{‘}s winning submissions, our multilingual system improved the translation quality on all language directions, with an average improvement of 2.0 BLEU. In the WMT2021 task, our system ranks first in 10 directions based on automatic evaluation.
@inproceedings{tran-etal-2021-facebook, title = "{F}acebook {AI}{'}s {WMT}21 News Translation Task Submission", author = "Tran, Chau and Bhosale, Shruti and Cross, James and Koehn, Philipp and Edunov, Sergey and Fan, Angela", editor = "Barrault, Loic and Bojar, Ondrej and Bougares, Fethi and Chatterjee, Rajen and Costa-jussa, Marta R. and Federmann, Christian and Fishel, Mark and Fraser, Alexander and Freitag, Markus and Graham, Yvette and Grundkiewicz, Roman and Guzman, Paco and Haddow, Barry and Huck, Matthias and Yepes, Antonio Jimeno and Koehn, Philipp and Kocmi, Tom and Martins, Andre and Morishita, Makoto and Monz, Christof", booktitle = "Proceedings of the Sixth Conference on Machine Translation", month = nov, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.wmt-1.19", pages = "205--215", abstract = "We describe Facebook{'}s multilingual model submission to the WMT2021 shared task on news translation. We participate in 14 language directions: English to and from Czech, German, Hausa, Icelandic, Japanese, Russian, and Chinese. To develop systems covering all these directions, we focus on multilingual models. We utilize data from all available sources {---} WMT, large-scale data mining, and in-domain backtranslation {---} to create high quality bilingual and multilingual baselines. Subsequently, we investigate strategies for scaling multilingual model size, such that one system has sufficient capacity for high quality representations of all eight languages. Our final submission is an ensemble of dense and sparse Mixture-of-Expert multilingual translation models, followed by finetuning on in-domain news data and noisy channel reranking. Compared to previous year{'}s winning submissions, our multilingual system improved the translation quality on all language directions, with an average improvement of 2.0 BLEU. In the WMT2021 task, our system ranks first in 10 directions based on automatic evaluation.", } - Jieneng Chen, Shuyang Sun, Ju He, Philip H. S. Torr, A. Yuille, and S. Bai, “TransMix: Attend to Mix for Vision Transformers,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{244346829, title = {TransMix: Attend to Mix for Vision Transformers}, author = {{Jieneng Chen} and {Shuyang Sun} and {Ju He} and {Philip H. S. Torr} and {A. Yuille} and {S. Bai}}, year = 2021, month = {11}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/b39495876b494412e0918898db8f988e9f5fd69d}, } - Zili Huang, Marc Delcroix, Leibny Paola García-Perera, Shinji Watanabe, Desh Raj, and S. Khudanpur, “Joint speaker diarization and speech recognition based on region proposal networks,” in Computer Speech and Language, 2021.
[BibTeX] [Link]@inproceedings{244107471, title = {Joint speaker diarization and speech recognition based on region proposal networks}, author = {{Zili Huang} and {Marc Delcroix} and {Leibny Paola García-Perera} and {Shinji Watanabe} and {Desh Raj} and {S. Khudanpur}}, year = 2021, month = {11}, booktitle = {Computer Speech and Language}, url = {https://www.semanticscholar.org/paper/9bb9b23823b45ba7521d872bb3e970ede4aafb8a}, } - T. Vieira, R. Cotterell, and J. Eisner, “Searching for More Efficient Dynamic Programs,” in Findings of EMNLP’21, Punta Cana, 2021, p. 3812–3830. doi:10.18653/v1/2021.findings-emnlp.322
[BibTeX] [Link]@InProceedings{vieira-et-al-2021-emnlp, aclid = "2021.findings-emnlp.322", doi = "10.18653/v1/2021.findings-emnlp.322", author = "Tim Vieira and Ryan Cotterell and Jason Eisner", title = "Searching for More Efficient Dynamic Programs", booktitle = "Findings of EMNLP'21", pages = "3812--3830", year = "2021", month = nov, address = "Punta Cana", URL = "http://cs.jhu.edu/~jason/papers/#vieira-et-al-2021-emnlp", } - R. Shin, C. H. Lin, S. Thomson, C. Chen, S. Roy, E. Antonios Platanios, A. Pauls, D. Klein, J. Eisner, and B. V. Durme, “Constrained Language Models Yield Few-Shot Semantic Parsers,” in Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Punta Cana, 2021, p. 7699–7715. doi:10.18653/v1/2021.emnlp-main.608
[BibTeX] [Link]@InProceedings{semanticmachines-2021-emnlp, aclid = "2021.emnlp-main.608", doi = "10.18653/v1/2021.emnlp-main.608", author = "Richard Shin and Christopher H. Lin and Sam Thomson and Charles Chen and Subhro Roy and Emmanouil Antonios Platanios and Adam Pauls and Dan Klein and Jason Eisner and Benjamin Van Durme", title = "Constrained Language Models Yield Few-Shot Semantic Parsers", booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing", pages = "7699--7715", year = "2021", month = nov, address = "Punta Cana", URL = "http://cs.jhu.edu/~jason/papers/#semanticmachines-2021-emnlp", } - K. Allen, Angeles Salles, Sa-Keun Park, Mounya Elhilali, and C. Moss, “Effect of background clutter on neural discrimination in the bat auditory midbrain.,” in Journal of Neurophysiology, 2021.
[BibTeX] [Link]@inproceedings{239051966, title = {Effect of background clutter on neural discrimination in the bat auditory midbrain.}, author = {{K. Allen} and {Angeles Salles} and {Sa-Keun Park} and {Mounya Elhilali} and {C. Moss}}, year = 2021, month = {10}, booktitle = {Journal of Neurophysiology}, url = {https://www.semanticscholar.org/paper/1652bdf2674f195b97aee0f1f32926f1c7b9aced}, } - Yixiao Zhang, Adam Kortylewski, Qing Liu, Seyoun Park, B. Green, E. Engle, Guillermo Almodovar, Ryan Walk, Sigfredo Soto-Diaz, J. Taube, A. Szalay, and A. Yuille, “A Light-Weight Interpretable Model for Nuclei Detection and Weakly-Supervised Segmentation,” in MOVI@MICCAI, 2021.
[BibTeX] [Link]@inproceedings{252440016, title = {A Light-Weight Interpretable Model for Nuclei Detection and Weakly-Supervised Segmentation}, author = {{Yixiao Zhang} and {Adam Kortylewski} and {Qing Liu} and {Seyoun Park} and {B. Green} and {E. Engle} and {Guillermo Almodovar} and {Ryan Walk} and {Sigfredo Soto-Diaz} and {J. Taube} and {A. Szalay} and {A. Yuille}}, year = 2021, month = {10}, booktitle = {MOVI@MICCAI}, url = {https://www.semanticscholar.org/paper/4795bf843f77bfd891e34729609c194b85b72a4d}, } - Shraman Pramanick, A. Roy, and Vishal M. Patel, “Multimodal Learning using Optimal Transport for Sarcasm and Humor Detection,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{239049720, title = {Multimodal Learning using Optimal Transport for Sarcasm and Humor Detection}, author = {{Shraman Pramanick} and {A. Roy} and {Vishal M. Patel}}, year = 2021, month = {10}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/204d5d9362533247df9a9303b44114c503236cdd}, } - Yu Zeng, Zhe L. Lin, Huchuan Lu, and Vishal M. Patel, “CR-Fill: Generative Image Inpainting with Auxiliary Contextual Reconstruction,” in IEEE International Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{244072324, title = {CR-Fill: Generative Image Inpainting with Auxiliary Contextual Reconstruction}, author = {{Yu Zeng} and {Zhe L. Lin} and {Huchuan Lu} and {Vishal M. Patel}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/2f1103a039c4511a111b506fdbe980a4f34b6709}, } - Wenpin Hou, Mingyu Zhang, Yuelong Ji, X. Hong, Guoying Wang, L. Liang, Hongkai Ji, S. Saria, and Xiaobin Wang, “In-Utero Exposure to Cigarette Smoking on Child Long-Term Risk of Obesity: Concordance of Self-Report, Maternal and Cord Blood Biomarkers.” 2021.
[BibTeX] [Link]@inproceedings{240189255, title = {In-Utero Exposure to Cigarette Smoking on Child Long-Term Risk of Obesity: Concordance of Self-Report, Maternal and Cord Blood Biomarkers}, author = {{Wenpin Hou} and {Mingyu Zhang} and {Yuelong Ji} and {X. Hong} and {Guoying Wang} and {L. Liang} and {Hongkai Ji} and {S. Saria} and {Xiaobin Wang}}, year = 2021, month = {10}, booktitle = {}, url = {https://www.semanticscholar.org/paper/eb17d81e0fdd641f07329cd202064e60db1aa2a3}, } - Aishan Liu, Xinyun Chen, Yingwei Li, Chaowei Xiao, Xun Yang, Xianglong Liu, D. Song, D. Tao, A. Yuille, and Anima Anandkumar, “ADVM’21: 1st International Workshop on Adversarial Learning for Multimedia,” in ACM Multimedia, 2021.
[BibTeX] [Link]@inproceedings{239011990, title = {ADVM'21: 1st International Workshop on Adversarial Learning for Multimedia}, author = {{Aishan Liu} and {Xinyun Chen} and {Yingwei Li} and {Chaowei Xiao} and {Xun Yang} and {Xianglong Liu} and {D. Song} and {D. Tao} and {A. Yuille} and {Anima Anandkumar}}, year = 2021, month = {10}, booktitle = {ACM Multimedia}, url = {https://www.semanticscholar.org/paper/943215bcb7866a6c6fe25944b14f41d5e2bd72b9}, } - Sandeep Reddy Kothinti, Nicholas Huang, and Mounya Elhilali, “Auditory salience using natural scenes: An online study,” in Journal of the Acoustical Society of America, 2021.
[BibTeX] [Link]@inproceedings{239454688, title = {Auditory salience using natural scenes: An online study}, author = {{Sandeep Reddy Kothinti} and {Nicholas Huang} and {Mounya Elhilali}}, year = 2021, month = {10}, booktitle = {Journal of the Acoustical Society of America}, url = {https://www.semanticscholar.org/paper/06ae11378419c01df4297c03d962459aefb3c054}, } - Matthew Wiesner, Desh Raj, and S. Khudanpur, “Injecting Text and Cross-Lingual Supervision in Few-Shot Learning from Self-Supervised Models,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2021.
[BibTeX] [Link]@inproceedings{238583266, title = {Injecting Text and Cross-Lingual Supervision in Few-Shot Learning from Self-Supervised Models}, author = {{Matthew Wiesner} and {Desh Raj} and {S. Khudanpur}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/047ce1b1f4dfec2d5f53de955f5e0f645ddc929c}, } - Piotr Żelasko, Daniel Povey, J. Trmal, and S. Khudanpur, “Lhotse: a speech data representation library for the modern deep learning ecosystem,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{239768221, title = {Lhotse: a speech data representation library for the modern deep learning ecosystem}, author = {{Piotr Żelasko} and {Daniel Povey} and {J. Trmal} and {S. Khudanpur}}, year = 2021, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/18394264fe8b4c05527117c5d15a1d19e52c2687}, } - M. Sophocleous, J. Georgiou, A. Andreou, Yosi Shacham-Diamand, Theerawit Wilaiprasitporn, J. Atkinson, Paddy J. French, E. García-Breijo, and Mohammad Russel, “Guest Editorial Special Issue on Sensors Tutorials: A Vigorous Dive Into the Vast Sea of Sensor- Related Knowledge—Part I,” in IEEE Sensors Journal, 2021.
[BibTeX] [Link]@inproceedings{245002248, title = {Guest Editorial Special Issue on Sensors Tutorials: A Vigorous Dive Into the Vast Sea of Sensor- Related Knowledge—Part I}, author = {{M. Sophocleous} and {J. Georgiou} and {A. Andreou} and {Yosi Shacham-Diamand} and {Theerawit Wilaiprasitporn} and {J. Atkinson} and {Paddy J. French} and {E. García-Breijo} and {Mohammad Russel}}, year = 2021, month = {10}, booktitle = {IEEE Sensors Journal}, url = {https://www.semanticscholar.org/paper/72e190cfe76cde934943ae35908bff346d4c970d}, } - Xinyue Wei, Weichao Qiu, Yi Zhang, Zihao Xiao, and A. Yuille, “Nuisance-Label Supervision: Robustness Improvement by Free Labels,” in 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), 2021.
[BibTeX] [Link]@inproceedings{238857299, title = {Nuisance-Label Supervision: Robustness Improvement by Free Labels}, author = {{Xinyue Wei} and {Weichao Qiu} and {Yi Zhang} and {Zihao Xiao} and {A. Yuille}}, year = 2021, month = {10}, booktitle = {2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)}, url = {https://www.semanticscholar.org/paper/0d8768aab838ec5c1af063fc95d22796fac05acf}, } - Angtian Wang, Shenxiao Mei, A. Yuille, and Adam Kortylewski, “Neural View Synthesis and Matching for Semi-Supervised Few-Shot Learning of 3D Pose,” in Neural Information Processing Systems, 2021.
[BibTeX] [Link]@inproceedings{239998658, title = {Neural View Synthesis and Matching for Semi-Supervised Few-Shot Learning of 3D Pose}, author = {{Angtian Wang} and {Shenxiao Mei} and {A. Yuille} and {Adam Kortylewski}}, year = 2021, month = {10}, booktitle = {Neural Information Processing Systems}, url = {https://www.semanticscholar.org/paper/f47d7c69997ba460133410eef2309be4eb29322c}, } - Shota Horiguchi, Yuki Takashima, Leibny Paola García-Perera, Shinji Watanabe, and Y. Kawaguchi, “Multi-Channel End-To-End Neural Diarization with Distributed Microphones,” in IEEE International Conference on Acoustics, Speech, and Signal Processing, 2021.
[BibTeX] [Link]@inproceedings{238583387, title = {Multi-Channel End-To-End Neural Diarization with Distributed Microphones}, author = {{Shota Horiguchi} and {Yuki Takashima} and {Leibny Paola García-Perera} and {Shinji Watanabe} and {Y. Kawaguchi}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Acoustics, Speech, and Signal Processing}, url = {https://www.semanticscholar.org/paper/04b44c518b145be625ff270af56cfd2e37900137}, } - Hossein Souri, Pirazh Khorramshahi, Chun Pong Lau, Micah Goldblum, and R. Chellappa, “Identification of Attack-Specific Signatures in Adversarial Examples,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{238743967, title = {Identification of Attack-Specific Signatures in Adversarial Examples}, author = {{Hossein Souri} and {Pirazh Khorramshahi} and {Chun Pong Lau} and {Micah Goldblum} and {R. Chellappa}}, year = 2021, month = {10}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7cfeca9f831e4f2d31114215abaa5078a98d1656}, } - VS Vibashan, Domenick Poster, Suya You, Shuowen Hu, and Vishal M. Patel, “Meta-UDA: Unsupervised Domain Adaptive Thermal Object Detection using Meta-Learning,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{238419143, title = {Meta-UDA: Unsupervised Domain Adaptive Thermal Object Detection using Meta-Learning}, author = {{VS Vibashan} and {Domenick Poster} and {Suya You} and {Shuowen Hu} and {Vishal M. Patel}}, year = 2021, month = {10}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/b9e3bd4e032adcdb4093a0cad5ae21d9eabbcee9}, } - Saurabhchand Bhati, J. Villalba, Piotr Żelasko, L. Moro-Velázquez, and N. Dehak, “Unsupervised Speech Segmentation and Variable Rate Representation Learning Using Segmental Contrastive Predictive Coding,” in IEEE/ACM Transactions on Audio Speech and Language Processing, 2021.
[BibTeX] [Link]@inproceedings{238408084, title = {Unsupervised Speech Segmentation and Variable Rate Representation Learning Using Segmental Contrastive Predictive Coding}, author = {{Saurabhchand Bhati} and {J. Villalba} and {Piotr Żelasko} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2021, month = {10}, booktitle = {IEEE/ACM Transactions on Audio Speech and Language Processing}, url = {https://www.semanticscholar.org/paper/3c2502b6d82ba4fca35fb871e7ed697fb4952f23}, } - Zhuowan Li, Elias Stengel-Eskin, Yixiao Zhang, Cihang Xie, Q. Tran, Benjamin Van Durme, and A. Yuille, “Calibrating Concepts and Operations: Towards Symbolic Reasoning on Real Images,” in IEEE International Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{238253118, title = {Calibrating Concepts and Operations: Towards Symbolic Reasoning on Real Images}, author = {{Zhuowan Li} and {Elias Stengel-Eskin} and {Yixiao Zhang} and {Cihang Xie} and {Q. Tran} and {Benjamin Van Durme} and {A. Yuille}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/40b065eb3aa5c5a54962aee78ebe30943beaabb1}, } - Rui Shao, Bochao Zhang, P. Yuen, and Vishal M. Patel, “Federated Test-Time Adaptive Face Presentation Attack Detection with Dual-Phase Privacy Preservation,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2021.
[BibTeX] [Link]@inproceedings{239768813, title = {Federated Test-Time Adaptive Face Presentation Attack Detection with Dual-Phase Privacy Preservation}, author = {{Rui Shao} and {Bochao Zhang} and {P. Yuen} and {Vishal M. Patel}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/3f3258ebf13c912d7de8df8a5a9446a702cd614c}, } - Joseph P. Robinson, Can Qin, Ming Shao, Matthew A. Turk, R. Chellappa, and Y. Fu, “The 5th Recognizing Families in the Wild Data Challenge: Predicting Kinship from Faces,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2021.
[BibTeX] [Link]@inproceedings{244728315, title = {The 5th Recognizing Families in the Wild Data Challenge: Predicting Kinship from Faces}, author = {{Joseph P. Robinson} and {Can Qin} and {Ming Shao} and {Matthew A. Turk} and {R. Chellappa} and {Y. Fu}}, year = 2021, month = {10}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/9f260bdd4030af5297a9c1cbb817c75701ac8c83}, } - W. Wu and D. Yarowsky, “On Pronunciations in Wiktionary: Extraction and Experiments on Multilingual Syllabification and Stress Prediction,” in Proceedings of the 14th Workshop on Building and Using Comparable Corpora (BUCC 2021), Online (Virtual Mode), 2021, p. 68–74.
[BibTeX] [Abstract] [Link]
We constructed parsers for five non-English editions of Wiktionary, which combined with pronunciations from the English edition, comprises over 5.3 million IPA pronunciations, the largest pronunciation lexicon of its kind. This dataset is a unique comparable corpus of IPA pronunciations annotated from multiple sources. We analyze the dataset, noting the presence of machine-generated pronunciations. We develop a novel visualization method to quantify syllabification. We experiment on the new combined task of multilingual IPA syllabification and stress prediction, finding that training a massively multilingual neural sequence-to-sequence model with copy attention can improve performance on both high- and low-resource languages, and multi-task training on stress prediction helps with syllabification.
@inproceedings{wu-yarowsky-2021-pronunciations, title = "On Pronunciations in {W}iktionary: Extraction and Experiments on Multilingual Syllabification and Stress Prediction", author = "Wu, Winston and Yarowsky, David", editor = "Rapp, Reinhard and Sharoff, Serge and Zweigenbaum, Pierre", booktitle = "Proceedings of the 14th Workshop on Building and Using Comparable Corpora (BUCC 2021)", month = sep, year = "2021", address = "Online (Virtual Mode)", publisher = "INCOMA Ltd.", url = "https://aclanthology.org/2021.bucc-1.9", pages = "68--74", abstract = "We constructed parsers for five non-English editions of Wiktionary, which combined with pronunciations from the English edition, comprises over 5.3 million IPA pronunciations, the largest pronunciation lexicon of its kind. This dataset is a unique comparable corpus of IPA pronunciations annotated from multiple sources. We analyze the dataset, noting the presence of machine-generated pronunciations. We develop a novel visualization method to quantify syllabification. We experiment on the new combined task of multilingual IPA syllabification and stress prediction, finding that training a massively multilingual neural sequence-to-sequence model with copy attention can improve performance on both high- and low-resource languages, and multi-task training on stress prediction helps with syllabification.", } - G. Kumar, P. Koehn, and S. Khudanpur, “Learning Curricula for Multilingual Neural Machine Translation Training,” in Proceedings of Machine Translation Summit XVIII: Research Track, Virtual, 2021, p. 1–9.
[BibTeX] [Abstract] [Link]
Low-resource Multilingual Neural Machine Translation (MNMT) is typically tasked with improving the translation performance on one or more language pairs with the aid of high-resource language pairs. In this paper and we propose two simple search based curricula {–} orderings of the multilingual training data {–} which help improve translation performance in conjunction with existing techniques such as fine-tuning. Additionally and we attempt to learn a curriculum for MNMT from scratch jointly with the training of the translation system using contextual multi-arm bandits. We show on the FLORES low-resource translation dataset that these learned curricula can provide better starting points for fine tuning and improve overall performance of the translation system.
@inproceedings{kumar-etal-2021-learning-curricula, title = "Learning Curricula for Multilingual Neural Machine Translation Training", author = "Kumar, Gaurav and Koehn, Philipp and Khudanpur, Sanjeev", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of Machine Translation Summit XVIII: Research Track", month = aug, year = "2021", address = "Virtual", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2021.mtsummit-research.1", pages = "1--9", abstract = "Low-resource Multilingual Neural Machine Translation (MNMT) is typically tasked with improving the translation performance on one or more language pairs with the aid of high-resource language pairs. In this paper and we propose two simple search based curricula {--} orderings of the multilingual training data {--} which help improve translation performance in conjunction with existing techniques such as fine-tuning. Additionally and we attempt to learn a curriculum for MNMT from scratch jointly with the training of the translation system using contextual multi-arm bandits. We show on the FLORES low-resource translation dataset that these learned curricula can provide better starting points for fine tuning and improve overall performance of the translation system.", } - H. Inaguma, B. Yan, S. Dalmia, P. Guo, J. Shi, K. Duh, and S. Watanabe, “ESPnet-ST IWSLT 2021 Offline Speech Translation System,” in Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021), Bangkok, Thailand (online), 2021, p. 100–109. doi:10.18653/v1/2021.iwslt-1.10
[BibTeX] [Abstract] [Link]
This paper describes the ESPnet-ST group{‘}s IWSLT 2021 submission in the offline speech translation track. This year we made various efforts on training data, architecture, and audio segmentation. On the data side, we investigated sequence-level knowledge distillation (SeqKD) for end-to-end (E2E) speech translation. Specifically, we used multi-referenced SeqKD from multiple teachers trained on different amounts of bitext. On the architecture side, we adopted the Conformer encoder and the Multi-Decoder architecture, which equips dedicated decoders for speech recognition and translation tasks in a unified encoder-decoder model and enables search in both source and target language spaces during inference. We also significantly improved audio segmentation by using the pyannote.audio toolkit and merging multiple short segments for long context modeling. Experimental evaluations showed that each of them contributed to large improvements in translation performance. Our best E2E system combined all the above techniques with model ensembling and achieved 31.4 BLEU on the 2-ref of tst2021 and 21.2 BLEU and 19.3 BLEU on the two single references of tst2021.
@inproceedings{inaguma-etal-2021-espnet, title = "{ESP}net-{ST} {IWSLT} 2021 Offline Speech Translation System", author = "Inaguma, Hirofumi and Yan, Brian and Dalmia, Siddharth and Guo, Pengcheng and Shi, Jiatong and Duh, Kevin and Watanabe, Shinji", editor = "Federico, Marcello and Waibel, Alex and Costa-juss{\`a}, Marta R. and Niehues, Jan and Stuker, Sebastian and Salesky, Elizabeth", booktitle = "Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021)", month = aug, year = "2021", address = "Bangkok, Thailand (online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.iwslt-1.10", doi = "10.18653/v1/2021.iwslt-1.10", pages = "100--109", abstract = "This paper describes the ESPnet-ST group{'}s IWSLT 2021 submission in the offline speech translation track. This year we made various efforts on training data, architecture, and audio segmentation. On the data side, we investigated sequence-level knowledge distillation (SeqKD) for end-to-end (E2E) speech translation. Specifically, we used multi-referenced SeqKD from multiple teachers trained on different amounts of bitext. On the architecture side, we adopted the Conformer encoder and the Multi-Decoder architecture, which equips dedicated decoders for speech recognition and translation tasks in a unified encoder-decoder model and enables search in both source and target language spaces during inference. We also significantly improved audio segmentation by using the pyannote.audio toolkit and merging multiple short segments for long context modeling. Experimental evaluations showed that each of them contributed to large improvements in translation performance. Our best E2E system combined all the above techniques with model ensembling and achieved 31.4 BLEU on the 2-ref of tst2021 and 21.2 BLEU and 19.3 BLEU on the two single references of tst2021.", } - R. Wicks and M. Post, “A unified approach to sentence segmentation of punctuated text in many languages,” in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), Online, 2021, p. 3995–4007. doi:10.18653/v1/2021.acl-long.309
[BibTeX] [Abstract] [Link]
The sentence is a fundamental unit of text processing. Yet sentences in the wild are commonly encountered not in isolation, but unsegmented within larger paragraphs and documents. Therefore, the first step in many NLP pipelines is \textit{sentence segmentation}. Despite its importance, this step is the subject of relatively little research. There are no standard test sets or even methods for evaluation, leaving researchers and engineers without a clear footing for evaluating and selecting models for the task. Existing tools have relatively small language coverage, and efforts to extend them to other languages are often ad hoc. We introduce a modern context-based modeling approach that provides a solution to the problem of segmenting punctuated text in many languages, and show how it can be trained on noisily-annotated data. We also establish a new 23-language multilingual evaluation set. Our approach exceeds high baselines set by existing methods on prior English corpora (WSJ and Brown corpora), and also performs well on average on our new evaluation set. We release our tool, ersatz, as open source.
@inproceedings{wicks-post-2021-unified, title = "A unified approach to sentence segmentation of punctuated text in many languages", author = "Wicks, Rachel and Post, Matt", editor = "Zong, Chengqing and Xia, Fei and Li, Wenjie and Navigli, Roberto", booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.acl-long.309", doi = "10.18653/v1/2021.acl-long.309", pages = "3995--4007", abstract = "The sentence is a fundamental unit of text processing. Yet sentences in the wild are commonly encountered not in isolation, but unsegmented within larger paragraphs and documents. Therefore, the first step in many NLP pipelines is \textit{sentence segmentation}. Despite its importance, this step is the subject of relatively little research. There are no standard test sets or even methods for evaluation, leaving researchers and engineers without a clear footing for evaluating and selecting models for the task. Existing tools have relatively small language coverage, and efforts to extend them to other languages are often ad hoc. We introduce a modern context-based modeling approach that provides a solution to the problem of segmenting punctuated text in many languages, and show how it can be trained on noisily-annotated data. We also establish a new 23-language multilingual evaluation set. Our approach exceeds high baselines set by existing methods on prior English corpora (WSJ and Brown corpora), and also performs well on average on our new evaluation set. We release our tool, ersatz, as open source.", } - W. Wu, K. Duh, and D. Yarowsky, “Sequence Models for Computational Etymology of Borrowings,” in Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, Online, 2021, p. 4032–4037. doi:10.18653/v1/2021.findings-acl.353
[BibTeX] [Link]@inproceedings{wu-etal-2021-sequence, title = "Sequence Models for Computational Etymology of Borrowings", author = "Wu, Winston and Duh, Kevin and Yarowsky, David", editor = "Zong, Chengqing and Xia, Fei and Li, Wenjie and Navigli, Roberto", booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.findings-acl.353", doi = "10.18653/v1/2021.findings-acl.353", pages = "4032--4037", } - L. Zhou, L. Ding, K. Duh, S. Watanabe, R. Sasano, and K. Takeda, “Self-Guided Curriculum Learning for Neural Machine Translation,” in Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021), Bangkok, Thailand (online), 2021, p. 206–214. doi:10.18653/v1/2021.iwslt-1.25
[BibTeX] [Abstract] [Link]
In supervised learning, a well-trained model should be able to recover ground truth accurately, i.e. the predicted labels are expected to resemble the ground truth labels as much as possible. Inspired by this, we formulate a difficulty criterion based on the recovery degrees of training examples. Motivated by the intuition that after skimming through the training corpus, the neural machine translation (NMT) model {“}knows{”} how to schedule a suitable curriculum according to learning difficulty, we propose a self-guided curriculum learning strategy that encourages the NMT model to learn from easy to hard on the basis of recovery degrees. Specifically, we adopt sentence-level BLEU score as the proxy of recovery degree. Experimental results on translation benchmarks including WMT14 English-German and WMT17 Chinese-English demonstrate that our proposed method considerably improves the recovery degree, thus consistently improving the translation performance.
@inproceedings{zhou-etal-2021-self, title = "Self-Guided Curriculum Learning for Neural Machine Translation", author = "Zhou, Lei and Ding, Liang and Duh, Kevin and Watanabe, Shinji and Sasano, Ryohei and Takeda, Koichi", editor = "Federico, Marcello and Waibel, Alex and Costa-juss{\`a}, Marta R. and Niehues, Jan and Stuker, Sebastian and Salesky, Elizabeth", booktitle = "Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021)", month = aug, year = "2021", address = "Bangkok, Thailand (online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.iwslt-1.25", doi = "10.18653/v1/2021.iwslt-1.25", pages = "206--214", abstract = "In supervised learning, a well-trained model should be able to recover ground truth accurately, i.e. the predicted labels are expected to resemble the ground truth labels as much as possible. Inspired by this, we formulate a difficulty criterion based on the recovery degrees of training examples. Motivated by the intuition that after skimming through the training corpus, the neural machine translation (NMT) model {``}knows{''} how to schedule a suitable curriculum according to learning difficulty, we propose a self-guided curriculum learning strategy that encourages the NMT model to learn from easy to hard on the basis of recovery degrees. Specifically, we adopt sentence-level BLEU score as the proxy of recovery degree. Experimental results on translation benchmarks including WMT14 English-German and WMT17 Chinese-English demonstrate that our proposed method considerably improves the recovery degree, thus consistently improving the translation performance.", } - N. Holzenberger and B. Van Durme, “Factoring Statutory Reasoning as Language Understanding Challenges,” in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), Online, 2021, p. 2742–2758. doi:10.18653/v1/2021.acl-long.213
[BibTeX] [Abstract] [Link]
Statutory reasoning is the task of determining whether a legal statute, stated in natural language, applies to the text description of a case. Prior work introduced a resource that approached statutory reasoning as a monolithic textual entailment problem, with neural baselines performing nearly at-chance. To address this challenge, we decompose statutory reasoning into four types of language-understanding challenge problems, through the introduction of concepts and structure found in Prolog programs. Augmenting an existing benchmark, we provide annotations for the four tasks, and baselines for three of them. Models for statutory reasoning are shown to benefit from the additional structure, improving on prior baselines. Further, the decomposition into subtasks facilitates finer-grained model diagnostics and clearer incremental progress.
@inproceedings{holzenberger-van-durme-2021-factoring, title = "Factoring Statutory Reasoning as Language Understanding Challenges", author = "Holzenberger, Nils and Van Durme, Benjamin", editor = "Zong, Chengqing and Xia, Fei and Li, Wenjie and Navigli, Roberto", booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.acl-long.213", doi = "10.18653/v1/2021.acl-long.213", pages = "2742--2758", abstract = "Statutory reasoning is the task of determining whether a legal statute, stated in natural language, applies to the text description of a case. Prior work introduced a resource that approached statutory reasoning as a monolithic textual entailment problem, with neural baselines performing nearly at-chance. To address this challenge, we decompose statutory reasoning into four types of language-understanding challenge problems, through the introduction of concepts and structure found in Prolog programs. Augmenting an existing benchmark, we provide annotations for the four tasks, and baselines for three of them. Models for statutory reasoning are shown to benefit from the additional structure, improving on prior baselines. Further, the decomposition into subtasks facilitates finer-grained model diagnostics and clearer incremental progress.", } - T. Q. Nguyen, K. Murray, and D. Chiang, “Data Augmentation by Concatenation for Low-Resource Translation: A Mystery and a Solution,” in Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021), Bangkok, Thailand (online), 2021, p. 287–293. doi:10.18653/v1/2021.iwslt-1.33
[BibTeX] [Abstract] [Link]
In this paper, we investigate the driving factors behind concatenation, a simple but effective data augmentation method for low-resource neural machine translation. Our experiments suggest that discourse context is unlikely the cause for concatenation improving BLEU by about +1 across four language pairs. Instead, we demonstrate that the improvement comes from three other factors unrelated to discourse: context diversity, length diversity, and (to a lesser extent) position shifting.
@inproceedings{nguyen-etal-2021-data, title = "Data Augmentation by Concatenation for Low-Resource Translation: A Mystery and a Solution", author = "Nguyen, Toan Q. and Murray, Kenton and Chiang, David", editor = "Federico, Marcello and Waibel, Alex and Costa-juss{\`a}, Marta R. and Niehues, Jan and Stuker, Sebastian and Salesky, Elizabeth", booktitle = "Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021)", month = aug, year = "2021", address = "Bangkok, Thailand (online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.iwslt-1.33", doi = "10.18653/v1/2021.iwslt-1.33", pages = "287--293", abstract = "In this paper, we investigate the driving factors behind concatenation, a simple but effective data augmentation method for low-resource neural machine translation. Our experiments suggest that discourse context is unlikely the cause for concatenation improving BLEU by about +1 across four language pairs. Instead, we demonstrate that the improvement comes from three other factors unrelated to discourse: context diversity, length diversity, and (to a lesser extent) position shifting.", } - J. Ou, N. Weir, A. Belyy, F. Yu, and B. Van Durme, “InFillmore: Frame-Guided Language Generation with Bidirectional Context,” in Proceedings of *SEM 2021: The Tenth Joint Conference on Lexical and Computational Semantics, Online, 2021, p. 129–142. doi:10.18653/v1/2021.starsem-1.12
[BibTeX] [Abstract] [Link]
We propose a structured extension to bidirectional-context conditional language generation, or {“}infilling,{”} inspired by Frame Semantic theory. Guidance is provided through one of two approaches: (1) model fine-tuning, conditioning directly on observed symbolic frames, and (2) a novel extension to disjunctive lexically constrained decoding that leverages frame semantic lexical units. Automatic and human evaluations confirm that frame-guided generation allows for explicit manipulation of intended infill semantics, with minimal loss in distinguishability from human-generated text. Our methods flexibly apply to a variety of use scenarios, and we provide an interactive web demo.
@inproceedings{ou-etal-2021-infillmore, title = "{I}n{F}illmore: Frame-Guided Language Generation with Bidirectional Context", author = "Ou, Jiefu and Weir, Nathaniel and Belyy, Anton and Yu, Felix and Van Durme, Benjamin", editor = "Ku, Lun-Wei and Nastase, Vivi and Vuli{\'c}, Ivan", booktitle = "Proceedings of *SEM 2021: The Tenth Joint Conference on Lexical and Computational Semantics", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.starsem-1.12", doi = "10.18653/v1/2021.starsem-1.12", pages = "129--142", abstract = "We propose a structured extension to bidirectional-context conditional language generation, or {``}infilling,{''} inspired by Frame Semantic theory. Guidance is provided through one of two approaches: (1) model fine-tuning, conditioning directly on observed symbolic frames, and (2) a novel extension to disjunctive lexically constrained decoding that leverages frame semantic lexical units. Automatic and human evaluations confirm that frame-guided generation allows for explicit manipulation of intended infill semantics, with minimal loss in distinguishability from human-generated text. Our methods flexibly apply to a variety of use scenarios, and we provide an interactive web demo.", } - X. Zhang and K. Duh, “Approaching Sign Language Gloss Translation as a Low-Resource Machine Translation Task,” in Proceedings of the 1st International Workshop on Automatic Translation for Signed and Spoken Languages (AT4SSL), Virtual, 2021, p. 60–70.
[BibTeX] [Abstract] [Link]
A cascaded Sign Language Translation system first maps sign videos to gloss annotations and then translates glosses into a spoken languages. This work focuses on the second-stage gloss translation component, which is challenging due to the scarcity of publicly available parallel data. We approach gloss translation as a low-resource machine translation task and investigate two popular methods for improving translation quality: hyperparameter search and backtranslation. We discuss the potentials and pitfalls of these methods based on experiments on the RWTH-PHOENIX-Weather 2014T dataset.
@inproceedings{zhang-duh-2021-approaching, title = "Approaching Sign Language Gloss Translation as a Low-Resource Machine Translation Task", author = "Zhang, Xuan and Duh, Kevin", editor = "Shterionov, Dimitar", booktitle = "Proceedings of the 1st International Workshop on Automatic Translation for Signed and Spoken Languages (AT4SSL)", month = aug, year = "2021", address = "Virtual", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2021.mtsummit-at4ssl.7", pages = "60--70", abstract = "A cascaded Sign Language Translation system first maps sign videos to gloss annotations and then translates glosses into a spoken languages. This work focuses on the second-stage gloss translation component, which is challenging due to the scarcity of publicly available parallel data. We approach gloss translation as a low-resource machine translation task and investigate two popular methods for improving translation quality: hyperparameter search and backtranslation. We discuss the potentials and pitfalls of these methods based on experiments on the RWTH-PHOENIX-Weather 2014T dataset.", } - K. Marchisio, P. Koehn, and C. Xiong, “An Alignment-Based Approach to Semi-Supervised Bilingual Lexicon Induction with Small Parallel Corpora,” in Proceedings of Machine Translation Summit XVIII: Research Track, Virtual, 2021, p. 293–304.
[BibTeX] [Abstract] [Link]
Aimed at generating a seed lexicon for use in downstream natural language tasks and unsupervised methods for bilingual lexicon induction have received much attention in the academic literature recently. While interesting and fully unsupervised settings are unrealistic; small amounts of bilingual data are usually available due to the existence of massively multilingual parallel corpora and or linguists can create small amounts of parallel data. In this work and we demonstrate an effective bootstrapping approach for semi-supervised bilingual lexicon induction that capitalizes upon the complementary strengths of two disparate methods for inducing bilingual lexicons. Whereas statistical methods are highly effective at inducing correct translation pairs for words frequently occurring in a parallel corpus and monolingual embedding spaces have the advantage of having been trained on large amounts of data and and therefore may induce accurate translations for words absent from the small corpus. By combining these relative strengths and our method achieves state-of-the-art results on 3 of 4 language pairs in the challenging VecMap test set using minimal amounts of parallel data and without the need for a translation dictionary. We release our implementation at www.blind-review.code.
@inproceedings{marchisio-etal-2021-alignment, title = "An Alignment-Based Approach to Semi-Supervised Bilingual Lexicon Induction with Small Parallel Corpora", author = "Marchisio, Kelly and Koehn, Philipp and Xiong, Conghao", editor = "Duh, Kevin and Guzm{\'a}n, Francisco", booktitle = "Proceedings of Machine Translation Summit XVIII: Research Track", month = aug, year = "2021", address = "Virtual", publisher = "Association for Machine Translation in the Americas", url = "https://aclanthology.org/2021.mtsummit-research.24", pages = "293--304", abstract = "Aimed at generating a seed lexicon for use in downstream natural language tasks and unsupervised methods for bilingual lexicon induction have received much attention in the academic literature recently. While interesting and fully unsupervised settings are unrealistic; small amounts of bilingual data are usually available due to the existence of massively multilingual parallel corpora and or linguists can create small amounts of parallel data. In this work and we demonstrate an effective bootstrapping approach for semi-supervised bilingual lexicon induction that capitalizes upon the complementary strengths of two disparate methods for inducing bilingual lexicons. Whereas statistical methods are highly effective at inducing correct translation pairs for words frequently occurring in a parallel corpus and monolingual embedding spaces have the advantage of having been trained on large amounts of data and and therefore may induce accurate translations for words absent from the small corpus. By combining these relative strengths and our method achieves state-of-the-art results on 3 of 4 language pairs in the challenging VecMap test set using minimal amounts of parallel data and without the need for a translation dictionary. We release our implementation at www.blind-review.code.", } - E. Schumacher, J. Mayfield, and M. Dredze, “Cross-Lingual Transfer in Zero-Shot Cross-Language Entity Linking,” in Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, Online, 2021, p. 583–595. doi:10.18653/v1/2021.findings-acl.52
[BibTeX] [Link]@inproceedings{schumacher-etal-2021-cross, title = "Cross-Lingual Transfer in Zero-Shot Cross-Language Entity Linking", author = "Schumacher, Elliot and Mayfield, James and Dredze, Mark", editor = "Zong, Chengqing and Xia, Fei and Li, Wenjie and Navigli, Roberto", booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.findings-acl.52", doi = "10.18653/v1/2021.findings-acl.52", pages = "583--595", } - G. Coppersmith, A. Fine, P. Crutchley, and J. Carroll, “Individual Differences in the Movement-Mood Relationship in Digital Life Data,” in Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access, Online, 2021, p. 25–31. doi:10.18653/v1/2021.clpsych-1.3
[BibTeX] [Abstract] [Link]
Our increasingly digitized lives generate troves of data that reflect our behavior, beliefs, mood, and wellbeing. Such {“}digital life data{”} provides crucial insight into the lives of patients outside the healthcare setting that has long been lacking, from a better understanding of mundane patterns of exercise and sleep routines to harbingers of emotional crisis. Moreover, information about individual differences and personalities is encoded in digital life data. In this paper we examine the relationship between mood and movement using linguistic and biometric data, respectively. Does increased physical activity (movement) have an effect on a person{‘}s mood (or vice-versa)? We find that weak group-level relationships between movement and mood mask interesting and often strong relationships between the two for individuals within the group. We describe these individual differences, and argue that individual variability in the relationship between movement and mood is one of many such factors that ought be taken into account in wellbeing-focused apps and AI systems.
@inproceedings{coppersmith-etal-2021-individual, title = "Individual Differences in the Movement-Mood Relationship in Digital Life Data", author = "Coppersmith, Glen and Fine, Alex and Crutchley, Patrick and Carroll, Joshua", editor = "Goharian, Nazli and Resnik, Philip and Yates, Andrew and Ireland, Molly and Niederhoffer, Kate and Resnik, Rebecca", booktitle = "Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access", month = jun, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.clpsych-1.3", doi = "10.18653/v1/2021.clpsych-1.3", pages = "25--31", abstract = "Our increasingly digitized lives generate troves of data that reflect our behavior, beliefs, mood, and wellbeing. Such {``}digital life data{''} provides crucial insight into the lives of patients outside the healthcare setting that has long been lacking, from a better understanding of mundane patterns of exercise and sleep routines to harbingers of emotional crisis. Moreover, information about individual differences and personalities is encoded in digital life data. In this paper we examine the relationship between mood and movement using linguistic and biometric data, respectively. Does increased physical activity (movement) have an effect on a person{'}s mood (or vice-versa)? We find that weak group-level relationships between movement and mood mask interesting and often strong relationships between the two for individuals within the group. We describe these individual differences, and argue that individual variability in the relationship between movement and mood is one of many such factors that ought be taken into account in wellbeing-focused apps and AI systems.", } - T. Lippincott and B. Van Durme, “Active learning and negative evidence for language identification,” in Proceedings of the Second Workshop on Data Science with Human in the Loop: Language Advances, Online, 2021, p. 47–51. doi:10.18653/v1/2021.dash-1.8
[BibTeX] [Abstract] [Link]
Language identification (LID), the task of determining the natural language of a given text, is an essential first step in most NLP pipelines. While generally a solved problem for documents of sufficient length and languages with ample training data, the proliferation of microblogs and other social media has made it increasingly common to encounter use-cases that *don{‘}t* satisfy these conditions. In these situations, the fundamental difficulty is the lack of, and cost of gathering, labeled data: unlike some annotation tasks, no single {“}expert{”} can quickly and reliably identify more than a handful of languages. This leads to a natural question: can we gain useful information when annotators are only able to *rule out* languages for a given document, rather than supply a positive label? What are the optimal choices for gathering and representing such *negative evidence* as a model is trained? In this paper, we demonstrate that using negative evidence can improve the performance of a simple neural LID model. This improvement is sensitive to policies of how the evidence is represented in the loss function, and for deciding which annotators to employ given the instance and model state. We consider simple policies and report experimental results that indicate the optimal choices for this task. We conclude with a discussion of future work to determine if and how the results generalize to other classification tasks.
@inproceedings{lippincott-van-durme-2021-active, title = "Active learning and negative evidence for language identification", author = "Lippincott, Thomas and Van Durme, Ben", editor = "Dragut, Eduard and Li, Yunyao and Popa, Lucian and Vucetic, Slobodan", booktitle = "Proceedings of the Second Workshop on Data Science with Human in the Loop: Language Advances", month = jun, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.dash-1.8", doi = "10.18653/v1/2021.dash-1.8", pages = "47--51", abstract = "Language identification (LID), the task of determining the natural language of a given text, is an essential first step in most NLP pipelines. While generally a solved problem for documents of sufficient length and languages with ample training data, the proliferation of microblogs and other social media has made it increasingly common to encounter use-cases that *don{'}t* satisfy these conditions. In these situations, the fundamental difficulty is the lack of, and cost of gathering, labeled data: unlike some annotation tasks, no single {``}expert{''} can quickly and reliably identify more than a handful of languages. This leads to a natural question: can we gain useful information when annotators are only able to *rule out* languages for a given document, rather than supply a positive label? What are the optimal choices for gathering and representing such *negative evidence* as a model is trained? In this paper, we demonstrate that using negative evidence can improve the performance of a simple neural LID model. This improvement is sensitive to policies of how the evidence is represented in the loss function, and for deciding which annotators to employ given the instance and model state. We consider simple policies and report experimental results that indicate the optimal choices for this task. We conclude with a discussion of future work to determine if and how the results generalize to other classification tasks.", } - S. MacAvaney, A. Mittu, G. Coppersmith, J. Leintz, and P. Resnik, “Community-level Research on Suicidality Prediction in a Secure Environment: Overview of the CLPsych 2021 Shared Task,” in Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access, Online, 2021, p. 70–80. doi:10.18653/v1/2021.clpsych-1.7
[BibTeX] [Abstract] [Link]
Progress on NLP for mental health {–-} indeed, for healthcare in general {–-} is hampered by obstacles to shared, community-level access to relevant data. We report on what is, to our knowledge, the first attempt to address this problem in mental health by conducting a shared task using sensitive data in a secure data enclave. Participating teams received access to Twitter posts donated for research, including data from users with and without suicide attempts, and did all work with the dataset entirely within a secure computational environment. We discuss the task, team results, and lessons learned to set the stage for future tasks on sensitive or confidential data.
@inproceedings{macavaney-etal-2021-community, title = "Community-level Research on Suicidality Prediction in a Secure Environment: Overview of the {CLP}sych 2021 Shared Task", author = "MacAvaney, Sean and Mittu, Anjali and Coppersmith, Glen and Leintz, Jeff and Resnik, Philip", editor = "Goharian, Nazli and Resnik, Philip and Yates, Andrew and Ireland, Molly and Niederhoffer, Kate and Resnik, Rebecca", booktitle = "Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access", month = jun, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.clpsych-1.7", doi = "10.18653/v1/2021.clpsych-1.7", pages = "70--80", abstract = "Progress on NLP for mental health {---} indeed, for healthcare in general {---} is hampered by obstacles to shared, community-level access to relevant data. We report on what is, to our knowledge, the first attempt to address this problem in mental health by conducting a shared task using sensitive data in a secure data enclave. Participating teams received access to Twitter posts donated for research, including data from users with and without suicide attempts, and did all work with the dataset entirely within a secure computational environment. We discuss the task, team results, and lessons learned to set the stage for future tasks on sensitive or confidential data.", } - Z. Wood-Doughty, P. Xu, X. Liu, and M. Dredze, “Using Noisy Self-Reports to Predict Twitter User Demographics,” in Proceedings of the Ninth International Workshop on Natural Language Processing for Social Media, Online, 2021, p. 123–137. doi:10.18653/v1/2021.socialnlp-1.11
[BibTeX] [Abstract] [Link]
Computational social science studies often contextualize content analysis within standard demographics. Since demographics are unavailable on many social media platforms (e.g. Twitter), numerous studies have inferred demographics automatically. Despite many studies presenting proof-of-concept inference of race and ethnicity, training of practical systems remains elusive since there are few annotated datasets. Existing datasets are small, inaccurate, or fail to cover the four most common racial and ethnic groups in the United States. We present a method to identify self-reports of race and ethnicity from Twitter profile descriptions. Despite the noise of automated supervision, our self-report datasets enable improvements in classification performance on gold standard self-report survey data. The result is a reproducible method for creating large-scale training resources for race and ethnicity.
@inproceedings{wood-doughty-etal-2021-using, title = "Using Noisy Self-Reports to Predict {T}witter User Demographics", author = "Wood-Doughty, Zach and Xu, Paiheng and Liu, Xiao and Dredze, Mark", editor = "Ku, Lun-Wei and Li, Cheng-Te", booktitle = "Proceedings of the Ninth International Workshop on Natural Language Processing for Social Media", month = jun, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.socialnlp-1.11", doi = "10.18653/v1/2021.socialnlp-1.11", pages = "123--137", abstract = "Computational social science studies often contextualize content analysis within standard demographics. Since demographics are unavailable on many social media platforms (e.g. Twitter), numerous studies have inferred demographics automatically. Despite many studies presenting proof-of-concept inference of race and ethnicity, training of practical systems remains elusive since there are few annotated datasets. Existing datasets are small, inaccurate, or fail to cover the four most common racial and ethnic groups in the United States. We present a method to identify self-reports of race and ethnicity from Twitter profile descriptions. Despite the noise of automated supervision, our self-report datasets enable improvements in classification performance on gold standard self-report survey data. The result is a reproducible method for creating large-scale training resources for race and ethnicity.", } - K. Harrigian, C. Aguirre, and M. Dredze, “On the State of Social Media Data for Mental Health Research,” in Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access, Online, 2021, p. 15–24. doi:10.18653/v1/2021.clpsych-1.2
[BibTeX] [Abstract] [Link]
Data-driven methods for mental health treatment and surveillance have become a major focus in computational science research in the last decade. However, progress in the domain remains bounded by the availability of adequate data. Prior systematic reviews have not necessarily made it possible to measure the degree to which data-related challenges have affected research progress. In this paper, we offer an analysis specifically on the state of social media data that exists for conducting mental health research. We do so by introducing an open-source directory of mental health datasets, annotated using a standardized schema to facilitate meta-analysis.
@inproceedings{harrigian-etal-2021-state, title = "On the State of Social Media Data for Mental Health Research", author = "Harrigian, Keith and Aguirre, Carlos and Dredze, Mark", editor = "Goharian, Nazli and Resnik, Philip and Yates, Andrew and Ireland, Molly and Niederhoffer, Kate and Resnik, Rebecca", booktitle = "Proceedings of the Seventh Workshop on Computational Linguistics and Clinical Psychology: Improving Access", month = jun, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.clpsych-1.2", doi = "10.18653/v1/2021.clpsych-1.2", pages = "15--24", abstract = "Data-driven methods for mental health treatment and surveillance have become a major focus in computational science research in the last decade. However, progress in the domain remains bounded by the availability of adequate data. Prior systematic reviews have not necessarily made it possible to measure the degree to which data-related challenges have affected research progress. In this paper, we offer an analysis specifically on the state of social media data that exists for conducting mental health research. We do so by introducing an open-source directory of mental health datasets, annotated using a standardized schema to facilitate meta-analysis.", } - C. Lin, A. Jaech, X. Li, Matt Gormley, and J. Eisner, “Limitations of Autoregressive Models and Their Alternatives,” in Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), Online, 2021, p. 5147–5173. doi:10.18653/v1/2021.naacl-main.405
[BibTeX] [Link]@InProceedings{lin-et-al-2021-naacl, aclid = "2021.naacl-main.405", doi = "10.18653/v1/2021.naacl-main.405", author = "Chu-Cheng Lin and Aaron Jaech and Xin Li and Matt Gormley and Jason Eisner", title = "Limitations of Autoregressive Models and Their Alternatives", booktitle = "Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)", pages = "5147--5173", year = "2021", month = jun, address = "Online", URL = "http://cs.jhu.edu/~jason/papers/#lin-et-al-2021-naacl", } - G. Qin and J. Eisner, “Learning How To Ask: Querying LMs with Mixtures of Soft Prompts,” in Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), Online, 2021, p. 5203–5212. doi:10.18653/v1/2021.naacl-main.410
[BibTeX] [Link]@InProceedings{qin-eisner-2021, aclid = "2021.naacl-main.410", doi = "10.18653/v1/2021.naacl-main.410", author = "Guanghui Qin and Jason Eisner", title = "Learning How To Ask: Querying {LM}s with Mixtures of Soft Prompts", booktitle = "Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)", pages = "5203--5212", year = "2021", month = jun, address = "Online", note = "Best Short Paper Award.", URL = "http://cs.jhu.edu/~jason/papers/#qin-eisner-2021", } - H. Xu, S. Ebner, M. Yarmohammadi, A. S. White, B. Van Durme, and K. Murray, “Gradual Fine-Tuning for Low-Resource Domain Adaptation,” in Proceedings of the Second Workshop on Domain Adaptation for NLP, Kyiv, Ukraine, 2021, p. 214–221.
[BibTeX] [Abstract] [Link]
Fine-tuning is known to improve NLP models by adapting an initial model trained on more plentiful but less domain-salient examples to data in a target domain. Such domain adaptation is typically done using one stage of fine-tuning. We demonstrate that gradually fine-tuning in a multi-step process can yield substantial further gains and can be applied without modifying the model or learning objective.
@inproceedings{xu-etal-2021-gradual, title = "Gradual Fine-Tuning for Low-Resource Domain Adaptation", author = "Xu, Haoran and Ebner, Seth and Yarmohammadi, Mahsa and White, Aaron Steven and Van Durme, Benjamin and Murray, Kenton", editor = "Ben-David, Eyal and Cohen, Shay and McDonald, Ryan and Plank, Barbara and Reichart, Roi and Rotman, Guy and Ziser, Yftah", booktitle = "Proceedings of the Second Workshop on Domain Adaptation for NLP", month = apr, year = "2021", address = "Kyiv, Ukraine", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.adaptnlp-1.22", pages = "214--221", abstract = "Fine-tuning is known to improve NLP models by adapting an initial model trained on more plentiful but less domain-salient examples to data in a target domain. Such domain adaptation is typically done using one stage of fine-tuning. We demonstrate that gradually fine-tuning in a multi-step process can yield substantial further gains and can be applied without modifying the model or learning objective.", } - J. Shi, J. D. Amith, R. Castillo Garc{‘i}a, E. Guadalupe Sierra, K. Duh, and S. Watanabe, “Leveraging End-to-End ASR for Endangered Language Documentation: An Empirical Study on Yolóxochitl Mixtec,” in Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, Online, 2021, p. 1134–1145. doi:10.18653/v1/2021.eacl-main.96
[BibTeX] [Abstract] [Link]
{“}Transcription bottlenecks{”}, created by a shortage of effective human transcribers (i.e., transcriber shortage), are one of the main challenges to endangered language (EL) documentation. Automatic speech recognition (ASR) has been suggested as a tool to overcome such bottlenecks. Following this suggestion, we investigated the effectiveness for EL documentation of end-to-end ASR, which unlike Hidden Markov Model ASR systems, eschews linguistic resources but is instead more dependent on large-data settings. We open source a Yoloxóchitl Mixtec EL corpus. First, we review our method in building an end-to-end ASR system in a way that would be reproducible by the ASR community. We then propose a novice transcription correction task and demonstrate how ASR systems and novice transcribers can work together to improve EL documentation. We believe this combinatory methodology would mitigate the transcription bottleneck and transcriber shortage that hinders EL documentation.
@inproceedings{shi-etal-2021-leveraging, title = "Leveraging End-to-End {ASR} for Endangered Language Documentation: An Empirical Study on Yol{\'o}xochitl {M}ixtec", author = "Shi, Jiatong and Amith, Jonathan D. and Castillo Garc{\'\i}a, Rey and Guadalupe Sierra, Esteban and Duh, Kevin and Watanabe, Shinji", editor = "Merlo, Paola and Tiedemann, Jorg and Tsarfaty, Reut", booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume", month = apr, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.eacl-main.96", doi = "10.18653/v1/2021.eacl-main.96", pages = "1134--1145", abstract = "{``}Transcription bottlenecks{''}, created by a shortage of effective human transcribers (i.e., transcriber shortage), are one of the main challenges to endangered language (EL) documentation. Automatic speech recognition (ASR) has been suggested as a tool to overcome such bottlenecks. Following this suggestion, we investigated the effectiveness for EL documentation of end-to-end ASR, which unlike Hidden Markov Model ASR systems, eschews linguistic resources but is instead more dependent on large-data settings. We open source a Yolox{\'o}chitl Mixtec EL corpus. First, we review our method in building an end-to-end ASR system in a way that would be reproducible by the ASR community. We then propose a novice transcription correction task and demonstrate how ASR systems and novice transcribers can work together to improve EL documentation. We believe this combinatory methodology would mitigate the transcription bottleneck and transcriber shortage that hinders EL documentation.", } - S. Sia and K. Duh, “Adaptive Mixed Component LDA for Low Resource Topic Modeling,” in Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, Online, 2021, p. 2451–2469. doi:10.18653/v1/2021.eacl-main.209
[BibTeX] [Abstract] [Link]
Probabilistic topic models in low data resource scenarios are faced with less reliable estimates due to sparsity of discrete word co-occurrence counts, and do not have the luxury of retraining word or topic embeddings using neural methods. In this challenging resource constrained setting, we explore mixture models which interpolate between the discrete and continuous topic-word distributions that utilise pre-trained embeddings to improve topic coherence. We introduce an automatic trade-off between the discrete and continuous representations via an adaptive mixture coefficient, which places greater weight on the discrete representation when the corpus statistics are more reliable. The adaptive mixture coefficient takes into account global corpus statistics, and the uncertainty in each topic{‘}s continuous distributions. Our approach outperforms the fully discrete, fully continuous, and static mixture model on topic coherence in low resource settings. We additionally demonstrate the generalisability of our method by extending it to handle multilingual document collections.
@inproceedings{sia-duh-2021-adaptive, title = "Adaptive Mixed Component {LDA} for Low Resource Topic Modeling", author = "Sia, Suzanna and Duh, Kevin", editor = "Merlo, Paola and Tiedemann, Jorg and Tsarfaty, Reut", booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume", month = apr, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.eacl-main.209", doi = "10.18653/v1/2021.eacl-main.209", pages = "2451--2469", abstract = "Probabilistic topic models in low data resource scenarios are faced with less reliable estimates due to sparsity of discrete word co-occurrence counts, and do not have the luxury of retraining word or topic embeddings using neural methods. In this challenging resource constrained setting, we explore mixture models which interpolate between the discrete and continuous topic-word distributions that utilise pre-trained embeddings to improve topic coherence. We introduce an automatic trade-off between the discrete and continuous representations via an adaptive mixture coefficient, which places greater weight on the discrete representation when the corpus statistics are more reliable. The adaptive mixture coefficient takes into account global corpus statistics, and the uncertainty in each topic{'}s continuous distributions. Our approach outperforms the fully discrete, fully continuous, and static mixture model on topic coherence in low resource settings. We additionally demonstrate the generalisability of our method by extending it to handle multilingual document collections.", } - X. Huang, M. J. Paul, F. Dernoncourt, R. Burke, and M. Dredze, “User Factor Adaptation for User Embedding via Multitask Learning,” in Proceedings of the Second Workshop on Domain Adaptation for NLP, Kyiv, Ukraine, 2021, p. 172–182.
[BibTeX] [Abstract] [Link]
Language varies across users and their interested fields in social media data: words authored by a user across his/her interests may have different meanings (e.g., cool) or sentiments (e.g., fast). However, most of the existing methods to train user embeddings ignore the variations across user interests, such as product and movie categories (e.g., drama vs. action). In this study, we treat the user interest as domains and empirically examine how the user language can vary across the user factor in three English social media datasets. We then propose a user embedding model to account for the language variability of user interests via a multitask learning framework. The model learns user language and its variations without human supervision. While existing work mainly evaluated the user embedding by extrinsic tasks, we propose an intrinsic evaluation via clustering and evaluate user embeddings by an extrinsic task, text classification. The experiments on the three English-language social media datasets show that our proposed approach can generally outperform baselines via adapting the user factor.
@inproceedings{huang-etal-2021-user, title = "User Factor Adaptation for User Embedding via Multitask Learning", author = "Huang, Xiaolei and Paul, Michael J. and Dernoncourt, Franck and Burke, Robin and Dredze, Mark", editor = "Ben-David, Eyal and Cohen, Shay and McDonald, Ryan and Plank, Barbara and Reichart, Roi and Rotman, Guy and Ziser, Yftah", booktitle = "Proceedings of the Second Workshop on Domain Adaptation for NLP", month = apr, year = "2021", address = "Kyiv, Ukraine", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.adaptnlp-1.18", pages = "172--182", abstract = "Language varies across users and their interested fields in social media data: words authored by a user across his/her interests may have different meanings (e.g., cool) or sentiments (e.g., fast). However, most of the existing methods to train user embeddings ignore the variations across user interests, such as product and movie categories (e.g., drama vs. action). In this study, we treat the user interest as domains and empirically examine how the user language can vary across the user factor in three English social media datasets. We then propose a user embedding model to account for the language variability of user interests via a multitask learning framework. The model learns user language and its variations without human supervision. While existing work mainly evaluated the user embedding by extrinsic tasks, we propose an intrinsic evaluation via clustering and evaluate user embeddings by an extrinsic task, text classification. The experiments on the three English-language social media datasets show that our proposed approach can generally outperform baselines via adapting the user factor.", } - M. Martindale, K. Duh, and M. Carpuat, “Machine Translation Believability,” in Proceedings of the First Workshop on Bridging Human–Computer Interaction and Natural Language Processing, Online, 2021, p. 88–95.
[BibTeX] [Abstract] [Link]
Successful Machine Translation (MT) deployment requires understanding not only the intrinsic qualities of MT output, such as fluency and adequacy, but also user perceptions. Users who do not understand the source language respond to MT output based on their perception of the likelihood that the meaning of the MT output matches the meaning of the source text. We refer to this as believability. Output that is not believable may be off-putting to users, but believable MT output with incorrect meaning may mislead them. In this work, we study the relationship of believability to fluency and adequacy by applying traditional MT direct assessment protocols to annotate all three features on the output of neural MT systems. Quantitative analysis of these annotations shows that believability is closely related to but distinct from fluency, and initial qualitative analysis suggests that semantic features may account for the difference.
@inproceedings{martindale-etal-2021-machine, title = "Machine Translation Believability", author = "Martindale, Marianna and Duh, Kevin and Carpuat, Marine", editor = "Blodgett, Su Lin and Madaio, Michael and O'Connor, Brendan and Wallach, Hanna and Yang, Qian", booktitle = "Proceedings of the First Workshop on Bridging Human{--}Computer Interaction and Natural Language Processing", month = apr, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.hcinlp-1.14", pages = "88--95", abstract = "Successful Machine Translation (MT) deployment requires understanding not only the intrinsic qualities of MT output, such as fluency and adequacy, but also user perceptions. Users who do not understand the source language respond to MT output based on their perception of the likelihood that the meaning of the MT output matches the meaning of the source text. We refer to this as believability. Output that is not believable may be off-putting to users, but believable MT output with incorrect meaning may mislead them. In this work, we study the relationship of believability to fluency and adequacy by applying traditional MT direct assessment protocols to annotate all three features on the output of neural MT systems. Quantitative analysis of these annotations shows that believability is closely related to but distinct from fluency, and initial qualitative analysis suggests that semantic features may account for the difference.", } - Anshul B. Shah, Shlok Kumar Mishra, Ankan Bansal, Jun-Cheng Chen, R. Chellappa, and Abhinav Shrivastava, “Pose and Joint-Aware Action Recognition-Supplementary Material.” 2021.
[BibTeX] [Link]@inproceedings{247112044, title = {Pose and Joint-Aware Action Recognition-Supplementary Material}, author = {{Anshul B. Shah} and {Shlok Kumar Mishra} and {Ankan Bansal} and {Jun-Cheng Chen} and {R. Chellappa} and {Abhinav Shrivastava}}, year = 2021, booktitle = {}, url = {https://www.semanticscholar.org/paper/7400177a4165c13d22da45a242ab8180e32a3d38}, } - Mintong Kang, Yongyi Lu, A. Yuille, and Zongwei Zhou, “Data, Assemble: Leveraging Multiple Datasets with Heterogeneous and Partial Labels,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{237940275, title = {Data, Assemble: Leveraging Multiple Datasets with Heterogeneous and Partial Labels}, author = {{Mintong Kang} and {Yongyi Lu} and {A. Yuille} and {Zongwei Zhou}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b269e5ae28f360b7ea159135a63ae1f82d9effbf}, } - E. Stengel-Eskin, J. Guallar-Blasco, and B. Van Durme, “Human-Model Divergence in the Handling of Vagueness,” in Proceedings of the 1st Workshop on Understanding Implicit and Underspecified Language, Online, 2021, p. 43–57. doi:10.18653/v1/2021.unimplicit-1.6
[BibTeX] [Abstract] [Link]
While aggregate performance metrics can generate valuable insights at a large scale, their dominance means more complex and nuanced language phenomena, such as vagueness, may be overlooked. Focusing on vague terms (e.g. sunny, cloudy, young, etc.) we inspect the behavior of visually grounded and text-only models, finding systematic divergences from human judgments even when a model{‘}s overall performance is high. To help explain this disparity, we identify two assumptions made by the datasets and models examined and, guided by the philosophy of vagueness, isolate cases where they do not hold.
@inproceedings{stengel-eskin-etal-2021-human, title = "Human-Model Divergence in the Handling of Vagueness", author = "Stengel-Eskin, Elias and Guallar-Blasco, Jimena and Van Durme, Benjamin", editor = "Roth, Michael and Tsarfaty, Reut and Goldberg, Yoav", booktitle = "Proceedings of the 1st Workshop on Understanding Implicit and Underspecified Language", month = aug, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.unimplicit-1.6", doi = "10.18653/v1/2021.unimplicit-1.6", pages = "43--57", abstract = "While aggregate performance metrics can generate valuable insights at a large scale, their dominance means more complex and nuanced language phenomena, such as vagueness, may be overlooked. Focusing on vague terms (e.g. sunny, cloudy, young, etc.) we inspect the behavior of visually grounded and text-only models, finding systematic divergences from human judgments even when a model{'}s overall performance is high. To help explain this disparity, we identify two assumptions made by the datasets and models examined and, guided by the philosophy of vagueness, isolate cases where they do not hold.", } - Matthew Maciejewski, Shinji Watanabe, and S. Khudanpur, “Speaker Verification-Based Evaluation of Single-Channel Speech Separation,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{239704028, title = {Speaker Verification-Based Evaluation of Single-Channel Speech Separation}, author = {{Matthew Maciejewski} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2021, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/39c5740304b5f4072f92e4e012a4b57e7bc2e817}, } - N. Balachandran, Jun-Cheng Chen, and R. Chellappa, “LR-to-HR Face Hallucination with an Adversarial Progressive Attribute-Induced Network,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{238226971, title = {LR-to-HR Face Hallucination with an Adversarial Progressive Attribute-Induced Network}, author = {{N. Balachandran} and {Jun-Cheng Chen} and {R. Chellappa}}, year = 2021, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/c741349663272c0d4a61e52d5650ba123bbbc81e}, } - Nicholas Ichien, Qing Liu, Shuhao Fu, K. Holyoak, A. Yuille, and Hongjing Lu, “Visual analogy: Deep learning versus compositional models,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{234742304, title = {Visual analogy: Deep learning versus compositional models}, author = {{Nicholas Ichien} and {Qing Liu} and {Shuhao Fu} and {K. Holyoak} and {A. Yuille} and {Hongjing Lu}}, year = 2021, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/1013e151d0aeeddac9e3c59db226c2ecf7d28c3f}, } - Sangwook Park, Woohyun Choi, and Mounya Elhilali, “Sound Event Detection with Cross-Referencing Self-Training.” 2021.
[BibTeX] [Link]@inproceedings{245930590, title = {Sound Event Detection with Cross-Referencing Self-Training}, author = {{Sangwook Park} and {Woohyun Choi} and {Mounya Elhilali}}, year = 2021, booktitle = {}, url = {https://www.semanticscholar.org/paper/14d161c0665ce7508c64270707ef62216b4e19a9}, } - R. Pappagari, Piotr Żelasko, J. Villalba, L. Moro-Velázquez, and N. Dehak, “Beyond Isolated Utterances: Conversational Emotion Recognition,” in Automatic Speech Recognition & Understanding, 2021.
[BibTeX] [Link]@inproceedings{237492280, title = {Beyond Isolated Utterances: Conversational Emotion Recognition}, author = {{R. Pappagari} and {Piotr Żelasko} and {J. Villalba} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2021, month = {9}, booktitle = {Automatic Speech Recognition & Understanding}, url = {https://www.semanticscholar.org/paper/6b39bd717627d97c7e69e46801fdbb38ef4eb946}, } - Domenick Poster, Matthew D. Thielke, R. Nguyen, Srinivasan Rajaraman, Xing Di, Cedric Nimpa Fondje, Vishal M. Patel, Nathan J. Short, B. Riggan, N. Nasrabadi, and Shuowen Hu, “A Large-Scale, Time-Synchronized Visible and Thermal Face Dataset,” in IEEE Workshop/Winter Conference on Applications of Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{230125224, title = {A Large-Scale, Time-Synchronized Visible and Thermal Face Dataset}, author = {{Domenick Poster} and {Matthew D. Thielke} and {R. Nguyen} and {Srinivasan Rajaraman} and {Xing Di} and {Cedric Nimpa Fondje} and {Vishal M. Patel} and {Nathan J. Short} and {B. Riggan} and {N. Nasrabadi} and {Shuowen Hu}}, year = 2021, month = {1}, booktitle = {IEEE Workshop/Winter Conference on Applications of Computer Vision}, url = {https://www.semanticscholar.org/paper/d88d4a05e076a070e1209245a40d57a0e9c211a2}, } - Desh Raj and S. Khudanpur, “Reformulating DOVER-Lap Label Mapping as a Graph Partitioning Problem,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{233025093, title = {Reformulating DOVER-Lap Label Mapping as a Graph Partitioning Problem}, author = {{Desh Raj} and {S. Khudanpur}}, year = 2021, month = {4}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/88dfc766aeff22a4e5fbdb81ce6161994c745039}, } - R. Yasarla, Federico Perazzi, and Vishal M. Patel, “Deblurring Face Images Using Deep Networks,” in Advances in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{238923351, title = {Deblurring Face Images Using Deep Networks}, author = {{R. Yasarla} and {Federico Perazzi} and {Vishal M. Patel}}, year = 2021, booktitle = {Advances in Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/8f28feea48abab5a40cb0926073fbb0cec4a77c9}, } - Pramuditha Perera, Poojan Oza, and Vishal M. Patel, “One-Class Classification: A Survey,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231418911, title = {One-Class Classification: A Survey}, author = {{Pramuditha Perera} and {Poojan Oza} and {Vishal M. Patel}}, year = 2021, month = {1}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/bd6262ebdd1a865e8e6859ab7dd8dc576d2a90e6}, } - A. Hussein, Shammur A. Chowdhury, N. Dehak, and Ahmed Ali, “Balanced End-to-End Monolingual pre-training for Low-Resourced Indic Languages Code-Switching Speech Recognition.” 2021.
[BibTeX] [Link]@inproceedings{246863876, title = {Balanced End-to-End Monolingual pre-training for Low-Resourced Indic Languages Code-Switching Speech Recognition}, author = {{A. Hussein} and {Shammur A. Chowdhury} and {N. Dehak} and {Ahmed Ali}}, year = 2021, month = {6}, booktitle = {}, url = {https://www.semanticscholar.org/paper/4781f897c02809c1522a06668ae1f4fa0e68e5ac}, } - Shao-Yuan Lo and Vishal M. Patel, “Error Diffusion Halftoning Against Adversarial Examples,” in International Conference on Information Photonics, 2021.
[BibTeX] [Link]@inproceedings{231698927, title = {Error Diffusion Halftoning Against Adversarial Examples}, author = {{Shao-Yuan Lo} and {Vishal M. Patel}}, year = 2021, month = {1}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/3b1fa137197c334b8929a326b607359e432cb68f}, } - Andrew Blair-Stanek and Benjamin Van Durme, “AI for Tax Analogies and Code Renumbering,” in Tax Law: Practitioner Series eJournal, 2021.
[BibTeX] [Link]@inproceedings{236512351, title = {AI for Tax Analogies and Code Renumbering}, author = {{Andrew Blair-Stanek} and {Benjamin Van Durme}}, year = 2021, month = {5}, booktitle = {Tax Law: Practitioner Series eJournal}, url = {https://www.semanticscholar.org/paper/a0c8218fcac4fe920b55bb930147b9aea58d3de2}, } - Chun Pong Lau, C. Castillo, and R. Chellappa, “ATFaceGAN: Single Face Semantic Aware Image Restoration and Recognition From Atmospheric Turbulence,” in IEEE Transactions on Biometrics Behavior and Identity Science, 2021.
[BibTeX] [Link]@inproceedings{232373611, title = {ATFaceGAN: Single Face Semantic Aware Image Restoration and Recognition From Atmospheric Turbulence}, author = {{Chun Pong Lau} and {C. Castillo} and {R. Chellappa}}, year = 2021, month = {4}, booktitle = {IEEE Transactions on Biometrics Behavior and Identity Science}, url = {https://www.semanticscholar.org/paper/d5ef84d04a6f527158d22304ff0bf73990d6563d}, } - Kevin Duh and Francisco Guzmán, “Proceedings of the 18th Biennial Machine Translation Summit (Volume 1: Research Track),” in Machine Translation Summit, 2021.
[BibTeX] [Link]@inproceedings{237206731, title = {Proceedings of the 18th Biennial Machine Translation Summit (Volume 1: Research Track)}, author = {{Kevin Duh} and {Francisco Guzmán}}, year = 2021, booktitle = {Machine Translation Summit}, url = {https://www.semanticscholar.org/paper/a693afc22d8cf7cbdf824a774c1c17195ae4c371}, } - Jieneng Chen, Yongyi Lu, Qihang Yu, Xiangde Luo, E. Adeli, Yan Wang, Le Lu, A. Yuille, and Yuyin Zhou, “TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231847326, title = {TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation}, author = {{Jieneng Chen} and {Yongyi Lu} and {Qihang Yu} and {Xiangde Luo} and {E. Adeli} and {Yan Wang} and {Le Lu} and {A. Yuille} and {Yuyin Zhou}}, year = 2021, month = {2}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/24b8a0b02bcb7934967757fc59d273a71ba67e30}, } - Ayah Zirikly, Bart Desmet, D. Newman-Griffis, Beth Marfeo, C. McDonough, H. Goldman, and L. Chan, “An Information Extraction Framework for Disability Determination: A mental Functioning Use-Case (Preprint).” 2021.
[BibTeX] [Link]@inproceedings{242764358, title = {An Information Extraction Framework for Disability Determination: A mental Functioning Use-Case (Preprint)}, author = {{Ayah Zirikly} and {Bart Desmet} and {D. Newman-Griffis} and {Beth Marfeo} and {C. McDonough} and {H. Goldman} and {L. Chan}}, year = 2021, month = {7}, booktitle = {}, url = {https://www.semanticscholar.org/paper/f068c7251b4a014fbe8f6e9cb722fd1c0f45da81}, } - Mintong Kang, Yongyi Lu, A. Yuille, and Zongwei Zhou, “Data-Assemble: Leveraging Multiple Datasets with Partial Labels.” 2021.
[BibTeX] [Link]@inproceedings{245007204, title = {Data-Assemble: Leveraging Multiple Datasets with Partial Labels}, author = {{Mintong Kang} and {Yongyi Lu} and {A. Yuille} and {Zongwei Zhou}}, year = 2021, booktitle = {}, url = {https://www.semanticscholar.org/paper/eeacac7a5f84274de89e185a975ca1f7584b858b}, } - D. Kelly, Max Spaderna, V. Hodzic, Glen A. Coppersmith, Shuo Chen, and P. Resnik, “Can language use in social media help in the treatment of severe mental illness?,” in Current research in psychiatry, 2021.
[BibTeX] [Link]@inproceedings{237533078, title = {Can language use in social media help in the treatment of severe mental illness?}, author = {{D. Kelly} and {Max Spaderna} and {V. Hodzic} and {Glen A. Coppersmith} and {Shuo Chen} and {P. Resnik}}, year = 2021, month = {8}, booktitle = {Current research in psychiatry}, url = {https://www.semanticscholar.org/paper/69ac548c8855aad23d897b52b0ba8d9bc4d8e107}, } - Ju He, Adam Kortylewski, and A. Yuille, “COMPAS: Representation Learning with Compositional Part Sharing for Few-Shot Classification,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231719786, title = {COMPAS: Representation Learning with Compositional Part Sharing for Few-Shot Classification}, author = {{Ju He} and {Adam Kortylewski} and {A. Yuille}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/3ad1caf62e0f4353ac3a7bb563392f7683999d23}, } - Chenxu Luo, Xiaodong Yang, and A. Yuille, “Exploring Simple 3D Multi-Object Tracking for Autonomous Driving,” in IEEE International Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{237266533, title = {Exploring Simple 3D Multi-Object Tracking for Autonomous Driving}, author = {{Chenxu Luo} and {Xiaodong Yang} and {A. Yuille}}, year = 2021, month = {8}, booktitle = {IEEE International Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/31115520c75bb9eb11ff2aee37c7605684d039f5}, } - Saurabh Kataria, J. Villalba, Piotr Żelasko, Laureano Moro-Vel’azquez, and N. Dehak, “Deep Feature CycleGANs: Speaker Identity Preserving Non-Parallel Microphone-Telephone Domain Adaptation for Speaker Verification,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{233024923, title = {Deep Feature CycleGANs: Speaker Identity Preserving Non-Parallel Microphone-Telephone Domain Adaptation for Speaker Verification}, author = {{Saurabh Kataria} and {J. Villalba} and {Piotr Żelasko} and {Laureano Moro-Vel'azquez} and {N. Dehak}}, year = 2021, month = {4}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/c3bb7ff3eba44535c9b704ee52041f91bde7bcd0}, } - Neehar Peri, Joshua Gleason, C. Castillo, T. Bourlai, Vishal M. Patel, and R. Chellappa, “A Synthesis-Based Approach for Thermal-to-Visible Face Verification,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2021.
[BibTeX] [Link]@inproceedings{237266437, title = {A Synthesis-Based Approach for Thermal-to-Visible Face Verification}, author = {{Neehar Peri} and {Joshua Gleason} and {C. Castillo} and {T. Bourlai} and {Vishal M. Patel} and {R. Chellappa}}, year = 2021, month = {8}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/edcfc2e222d08c51a9f1087fb29252b659d9b071}, } - Yuki Takashima, Yusuke Fujita, Shinji Watanabe, Shota Horiguchi, Leibny Paola García-Perera, and Kenji Nagamatsu, “End-to-End Speaker Diarization Conditioned on Speech Activity and Overlap Detection,” in Spoken Language Technology Workshop, 2021.
[BibTeX] [Link]@inproceedings{232413801, title = {End-to-End Speaker Diarization Conditioned on Speech Activity and Overlap Detection}, author = {{Yuki Takashima} and {Yusuke Fujita} and {Shinji Watanabe} and {Shota Horiguchi} and {Leibny Paola García-Perera} and {Kenji Nagamatsu}}, year = 2021, month = {1}, booktitle = {Spoken Language Technology Workshop}, url = {https://www.semanticscholar.org/paper/cbf9a2560eac548e7b3d5eb7074c40b7bb861909}, } - M. Illa, B. Halpern, R. V. Son, L. Moro-Velázquez, and O. Scharenborg, “Pathological voice adaptation with autoencoder-based voice conversion,” in 11th ISCA Speech Synthesis Workshop (SSW 11), 2021.
[BibTeX] [Link]@inproceedings{235446502, title = {Pathological voice adaptation with autoencoder-based voice conversion}, author = {{M. Illa} and {B. Halpern} and {R. V. Son} and {L. Moro-Velázquez} and {O. Scharenborg}}, year = 2021, month = {6}, booktitle = {11th ISCA Speech Synthesis Workshop (SSW 11)}, url = {https://www.semanticscholar.org/paper/28519a0c7bf3de835b184f3c85f01f4d1d8746d8}, } - VS Vibashan, Jeya Maria Jose Valanarasu, Poojan Oza, and Vishal M. Patel, “Image Fusion Transformer,” in International Conference on Information Photonics, 2021.
[BibTeX] [Link]@inproceedings{236087620, title = {Image Fusion Transformer}, author = {{VS Vibashan} and {Jeya Maria Jose Valanarasu} and {Poojan Oza} and {Vishal M. Patel}}, year = 2021, month = {7}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/48ec7f1bcf8953ac472384bcea88bc38774508f0}, } - Tiange Xiang, Yongyi Lu, A. Yuille, Chaoyi Zhang, Weidong (Tom) Cai, and Zongwei Zhou, “In-painting Radiography Images for Unsupervised Anomaly Detection,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{244709610, title = {In-painting Radiography Images for Unsupervised Anomaly Detection}, author = {{Tiange Xiang} and {Yongyi Lu} and {A. Yuille} and {Chaoyi Zhang} and {Weidong (Tom) Cai} and {Zongwei Zhou}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/0a0adec915e0a0165e9d048df46dc11404d8f9ca}, } - Kelly Marchisio, Youngser Park, Ali Saad-Eldin, A. Alyakin, Kevin Duh, C. Priebe, and Philipp Koehn, “An Analysis of Euclidean vs. Graph-Based Framing for Bilingual Lexicon Induction from Word Embedding Spaces,” in Conference on Empirical Methods in Natural Language Processing, 2021.
[BibTeX] [Link]@inproceedings{237941142, title = {An Analysis of Euclidean vs. Graph-Based Framing for Bilingual Lexicon Induction from Word Embedding Spaces}, author = {{Kelly Marchisio} and {Youngser Park} and {Ali Saad-Eldin} and {A. Alyakin} and {Kevin Duh} and {C. Priebe} and {Philipp Koehn}}, year = 2021, month = {9}, booktitle = {Conference on Empirical Methods in Natural Language Processing}, url = {https://www.semanticscholar.org/paper/0a5fc6d1735dd2761fc31fad5a3b40a4fa06546b}, } - Navaneeth Bodla, G. Shrivastava, R. Chellappa, and Abhinav Shrivastava, “Hierarchical Video Prediction using Relational Layouts for Human-Object Interactions,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{235691818, title = {Hierarchical Video Prediction using Relational Layouts for Human-Object Interactions}, author = {{Navaneeth Bodla} and {G. Shrivastava} and {R. Chellappa} and {Abhinav Shrivastava}}, year = 2021, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/302e4537b277384542d7f0b5cdc4db33abbaa1db}, } - Angtian Wang, Adam Kortylewski, and A. Yuille, “NeMo: Neural Mesh Models of Contrastive Features for Robust 3D Pose Estimation,” in International Conference on Learning Representations, 2021.
[BibTeX] [Link]@inproceedings{231728364, title = {NeMo: Neural Mesh Models of Contrastive Features for Robust 3D Pose Estimation}, author = {{Angtian Wang} and {Adam Kortylewski} and {A. Yuille}}, year = 2021, month = {1}, booktitle = {International Conference on Learning Representations}, url = {https://www.semanticscholar.org/paper/74c61e2d2f12e56015741bcc2340f14a776e8606}, } - Zach Wood-Doughty, I. Shpitser, and Mark Dredze, “Generating Synthetic Text Data to Evaluate Causal Inference Methods,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231861828, title = {Generating Synthetic Text Data to Evaluate Causal Inference Methods}, author = {{Zach Wood-Doughty} and {I. Shpitser} and {Mark Dredze}}, year = 2021, month = {2}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/9adc1a3307c05ff3c9b0ae595cb57b1de041713f}, } - Shota Horiguchi, Yusuke Fujita, Shinji Watanabe, Yawen Xue, and Leibny Paola García-Perera, “Encoder-Decoder Based Attractor Calculation for End-to-End Neural Diarization,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{235489753, title = {Encoder-Decoder Based Attractor Calculation for End-to-End Neural Diarization}, author = {{Shota Horiguchi} and {Yusuke Fujita} and {Shinji Watanabe} and {Yawen Xue} and {Leibny Paola García-Perera}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/8abd724b770348bd21b16b9aaf2ba0a77596b2ed}, } - Kelly Marchisio, Conghao Xiong, and Philipp Koehn, “Embedding-Enhanced Giza++: Improving Alignment in Low- and High- Resource Scenarios Using Embedding Space Geometry,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{260435370, title = {Embedding-Enhanced Giza++: Improving Alignment in Low- and High- Resource Scenarios Using Embedding Space Geometry}, author = {{Kelly Marchisio} and {Conghao Xiong} and {Philipp Koehn}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/f7bb9e4c44356182bff28bfca99404c7d9c04d2b}, } - Pramuditha Perera and Vishal M. Patel, “Geometric Transformation-Based Network Ensemble for Open-Set Recognition,” in IEEE International Conference on Multimedia and Expo, 2021.
[BibTeX] [Link]@inproceedings{236264069, title = {Geometric Transformation-Based Network Ensemble for Open-Set Recognition}, author = {{Pramuditha Perera} and {Vishal M. Patel}}, year = 2021, month = {7}, booktitle = {IEEE International Conference on Multimedia and Expo}, url = {https://www.semanticscholar.org/paper/6ef59ba79e3c4d725bfdd22174f2adc72d801245}, } - Pengfei Guo, Jeya Maria Jose Valanarasu, Puyang Wang, Jinyuan Zhou, Shanshan Jiang, and Vishal M. Patel, “Over-and-Under Complete Convolutional RNN for MRI Reconstruction,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2021.
[BibTeX] [Link]@inproceedings{235446931, title = {Over-and-Under Complete Convolutional RNN for MRI Reconstruction}, author = {{Pengfei Guo} and {Jeya Maria Jose Valanarasu} and {Puyang Wang} and {Jinyuan Zhou} and {Shanshan Jiang} and {Vishal M. Patel}}, year = 2021, month = {6}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/4e5095ca6e280b068aa572c6d4afc32d6b246492}, } - Nithin Gopalakrishnan Nair and Vishal M. Patel, “Confidence Guided Network For Atmospheric Turbulence Mitigation,” in International Conference on Information Photonics, 2021.
[BibTeX] [Link]@inproceedings{238681785, title = {Confidence Guided Network For Atmospheric Turbulence Mitigation}, author = {{Nithin Gopalakrishnan Nair} and {Vishal M. Patel}}, year = 2021, month = {9}, booktitle = {International Conference on Information Photonics}, url = {https://www.semanticscholar.org/paper/34ff864bcef1d3f8bbacc3c241ee65cc6b13b84e}, } - Guoguo Chen, Shuzhou Chai, Guan-Bo Wang, Jiayu Du, Weiqiang Zhang, Chao Weng, Dan Su, Daniel Povey, J. Trmal, Junbo Zhang, Mingjie Jin, S. Khudanpur, Shinji Watanabe, Shuaijiang Zhao, Wei Zou, Xiangang Li, Xuchen Yao, Yongqing Wang, Yujun Wang, Zhao You, and Zhiyong Yan, “GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10, 000 Hours of Transcribed Audio,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{235422086, title = {GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10, 000 Hours of Transcribed Audio}, author = {{Guoguo Chen} and {Shuzhou Chai} and {Guan-Bo Wang} and {Jiayu Du} and {Weiqiang Zhang} and {Chao Weng} and {Dan Su} and {Daniel Povey} and {J. Trmal} and {Junbo Zhang} and {Mingjie Jin} and {S. Khudanpur} and {Shinji Watanabe} and {Shuaijiang Zhao} and {Wei Zou} and {Xiangang Li} and {Xuchen Yao} and {Yongqing Wang} and {Yujun Wang} and {Zhao You} and {Zhiyong Yan}}, year = 2021, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/6f1ca0249eafa36a5762ac53f6ba2a4ee2133456}, } - Jonah P. Sengupta, M. Villemur, Daniel R. Mendat, Gaspar Tognetti, and A. Andreou, “Architecture and Algorithm Co-Design Framework for Embedded Processors in Event-Based Cameras,” in International Symposium on Circuits and Systems, 2021.
[BibTeX] [Link]@inproceedings{235520107, title = {Architecture and Algorithm Co-Design Framework for Embedded Processors in Event-Based Cameras}, author = {{Jonah P. Sengupta} and {M. Villemur} and {Daniel R. Mendat} and {Gaspar Tognetti} and {A. Andreou}}, year = 2021, month = {5}, booktitle = {International Symposium on Circuits and Systems}, url = {https://www.semanticscholar.org/paper/c7bc38e1a275d8e17aa779f0d66c567398c5d0cb}, } - S. Saria, Peter F. Schulam, Brian J. Yeh, Daniel Burke, S. Mooney, C. Fong, Jacob E. Sunshine, Dustin R. Long, and V. O’Reilly-Shah, “Development and Validation of ARC, a Model for Anticipating Acute Respiratory Failure in Coronavirus Disease 2019 Patients,” in Critical Care Explorations, 2021.
[BibTeX] [Link]@inproceedings{235371744, title = {Development and Validation of ARC, a Model for Anticipating Acute Respiratory Failure in Coronavirus Disease 2019 Patients}, author = {{S. Saria} and {Peter F. Schulam} and {Brian J. Yeh} and {Daniel Burke} and {S. Mooney} and {C. Fong} and {Jacob E. Sunshine} and {Dustin R. Long} and {V. O’Reilly-Shah}}, year = 2021, month = {6}, booktitle = {Critical Care Explorations}, url = {https://www.semanticscholar.org/paper/947fad4b54ebf666dc5c73837beb9bf3e018e7c6}, } - M. Landers, R. Dorsey, and S. Saria, “Digital Endpoints: Definition, Benefits, and Current Barriers in Accelerating Development and Adoption,” in Digital Biomarkers, 2021.
[BibTeX] [Link]@inproceedings{239134009, title = {Digital Endpoints: Definition, Benefits, and Current Barriers in Accelerating Development and Adoption}, author = {{M. Landers} and {R. Dorsey} and {S. Saria}}, year = 2021, month = {9}, booktitle = {Digital Biomarkers}, url = {https://www.semanticscholar.org/paper/a79a3ea8141ac2d9f780abbcd1ee0b2bfbd78ead}, } - Elizabeth Salesky, Matthew Wiesner, Jacob Bremerman, R. Cattoni, Matteo Negri, M. Turchi, Douglas W. Oard, and Matt Post, “The Multilingual TEDx Corpus for Speech Recognition and Translation,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{231786401, title = {The Multilingual TEDx Corpus for Speech Recognition and Translation}, author = {{Elizabeth Salesky} and {Matthew Wiesner} and {Jacob Bremerman} and {R. Cattoni} and {Matteo Negri} and {M. Turchi} and {Douglas W. Oard} and {Matt Post}}, year = 2021, month = {2}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/8e1e7741b56455056ff369fff9889b4c5f998b58}, } - Chen Wei, Kihyuk Sohn, Clayton Mellina, A. Yuille, and Fan Yang, “CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{231951327, title = {CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning}, author = {{Chen Wei} and {Kihyuk Sohn} and {Clayton Mellina} and {A. Yuille} and {Fan Yang}}, year = 2021, month = {2}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/09acced5fcb49322f5a26ac7a4cbe9f1308657c4}, } - Noah Weber, Anton Belyy, Nils Holzenberger, Rachel Rudinger, and Benjamin Van Durme, “Schema Curation via Causal Association Rule Mining,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{233296523, title = {Schema Curation via Causal Association Rule Mining}, author = {{Noah Weber} and {Anton Belyy} and {Nils Holzenberger} and {Rachel Rudinger} and {Benjamin Van Durme}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/1da3dc6f997599a197b1510f2c2bdd86e2d86e49}, } - Yunjuan Wang, Poorya Mianjy, and R. Arora, “Robust Learning for Data Poisoning Attacks,” in International Conference on Machine Learning, 2021.
[BibTeX] [Link]@inproceedings{235826166, title = {Robust Learning for Data Poisoning Attacks}, author = {{Yunjuan Wang} and {Poorya Mianjy} and {R. Arora}}, year = 2021, booktitle = {International Conference on Machine Learning}, url = {https://www.semanticscholar.org/paper/c541fa104bc5297f3ebf967855d582ab9a37291d}, } - I. McLane, Dimitra Emmanouilidou, James E. West, and Mounya Elhilali, “Design and Comparative Performance of a Robust Lung Auscultation System for Noisy Clinical Settings,” in IEEE journal of biomedical and health informatics, 2021.
[BibTeX] [Link]@inproceedings{231805407, title = {Design and Comparative Performance of a Robust Lung Auscultation System for Noisy Clinical Settings}, author = {{I. McLane} and {Dimitra Emmanouilidou} and {James E. West} and {Mounya Elhilali}}, year = 2021, month = {2}, booktitle = {IEEE journal of biomedical and health informatics}, url = {https://www.semanticscholar.org/paper/84d283da84a56296c925a6c383bad6e4cb345376}, } - H. Inaguma, Yosuke Higuchi, Kevin Duh, Tatsuya Kawahara, and Shinji Watanabe, “Non-autoregressive End-to-end Speech Translation with Parallel Autoregressive Rescoring,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{237453587, title = {Non-autoregressive End-to-end Speech Translation with Parallel Autoregressive Rescoring}, author = {{H. Inaguma} and {Yosuke Higuchi} and {Kevin Duh} and {Tatsuya Kawahara} and {Shinji Watanabe}}, year = 2021, month = {9}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/d79b613a67cf79740e1c08037f7d054585a12284}, } - Saurabhchand Bhati, J. Villalba, Piotr Żelasko, L. Moro-Velázquez, and N. Dehak, “Segmental Contrastive Predictive Coding for Unsupervised Word Segmentation,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{235352541, title = {Segmental Contrastive Predictive Coding for Unsupervised Word Segmentation}, author = {{Saurabhchand Bhati} and {J. Villalba} and {Piotr Żelasko} and {L. Moro-Velázquez} and {N. Dehak}}, year = 2021, month = {6}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/642dab29e680f516eb25949d616a24e0ad147a19}, } - Jeya Maria Jose Valanarasu, Poojan Oza, I. Hacihaliloglu, and Vishal M. Patel, “Medical Transformer: Gated Axial-Attention for Medical Image Segmentation,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, 2021.
[BibTeX] [Link]@inproceedings{231986084, title = {Medical Transformer: Gated Axial-Attention for Medical Image Segmentation}, author = {{Jeya Maria Jose Valanarasu} and {Poojan Oza} and {I. Hacihaliloglu} and {Vishal M. Patel}}, year = 2021, month = {2}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, url = {https://www.semanticscholar.org/paper/1e5e8106700c8dbdfa036a5a9be5e61e06c0ed02}, } - Shota Horiguchi, Nelson Yalta, Leibny Paola García-Perera, Yuki Takashima, Yawen Xue, Desh Raj, Zili Huang, Yusuke Fujita, Shinji Watanabe, and S. Khudanpur, “The Hitachi-JHU DIHARD III System: Competitive End-to-End Neural Diarization and X-Vector Clustering Systems Combined by DOVER-Lap,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231749872, title = {The Hitachi-JHU DIHARD III System: Competitive End-to-End Neural Diarization and X-Vector Clustering Systems Combined by DOVER-Lap}, author = {{Shota Horiguchi} and {Nelson Yalta} and {Leibny Paola García-Perera} and {Yuki Takashima} and {Yawen Xue} and {Desh Raj} and {Zili Huang} and {Yusuke Fujita} and {Shinji Watanabe} and {S. Khudanpur}}, year = 2021, month = {2}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7a737872a6693ba3f0c99651191b93dad0dadcee}, } - Jonah P. Sengupta, M. Villemur, and A. Andreou, “A Spike-based Cellular-Neural Network Architecture for Spatiotemporal filtering,” in Annual Conference on Information Sciences and Systems, 2021.
[BibTeX] [Link]@inproceedings{233333189, title = {A Spike-based Cellular-Neural Network Architecture for Spatiotemporal filtering}, author = {{Jonah P. Sengupta} and {M. Villemur} and {A. Andreou}}, year = 2021, month = {3}, booktitle = {Annual Conference on Information Sciences and Systems}, url = {https://www.semanticscholar.org/paper/76791fe786d8fd412ee15ca19b65c8e5b3103bc1}, } - Qing Liu, Adam Kortylewski, Zhishuai Zhang, Zizhang Li, Mengqi Guo, Qihao Liu, Xiaoding Yuan, Jiteng Mu, Weichao Qiu, and A. Yuille, “Learning Part Segmentation through Unsupervised Domain Adaptation from Synthetic Vehicles,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{247939729, title = {Learning Part Segmentation through Unsupervised Domain Adaptation from Synthetic Vehicles}, author = {{Qing Liu} and {Adam Kortylewski} and {Zhishuai Zhang} and {Zizhang Li} and {Mengqi Guo} and {Qihao Liu} and {Xiaoding Yuan} and {Jiteng Mu} and {Weichao Qiu} and {A. Yuille}}, year = 2021, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/f67fe2f05ccbfa7eb45fe0f8ed99e2be4279e3e7}, } - P. Xia, G. Qin, S. Vashishtha, Y. Chen, T. Chen, C. May, C. Harman, K. Rawlins, A. S. White, and B. Van Durme, “LOME: Large Ontology Multilingual Extraction,” in Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, Online, 2021, p. 149–159. doi:10.18653/v1/2021.eacl-demos.19
[BibTeX] [Abstract] [Link]
We present LOME, a system for performing multilingual information extraction. Given a text document as input, our core system identifies spans of textual entity and event mentions with a FrameNet (Baker et al., 1998) parser. It subsequently performs coreference resolution, fine-grained entity typing, and temporal relation prediction between events. By doing so, the system constructs an event and entity focused knowledge graph. We can further apply third-party modules for other types of annotation, like relation extraction. Our (multilingual) first-party modules either outperform or are competitive with the (monolingual) state-of-the-art. We achieve this through the use of multilingual encoders like XLM-R (Conneau et al., 2020) and leveraging multilingual training data. LOME is available as a Docker container on Docker Hub. In addition, a lightweight version of the system is accessible as a web demo.
@inproceedings{xia-etal-2021-lome, title = "{LOME}: Large Ontology Multilingual Extraction", author = "Xia, Patrick and Qin, Guanghui and Vashishtha, Siddharth and Chen, Yunmo and Chen, Tongfei and May, Chandler and Harman, Craig and Rawlins, Kyle and White, Aaron Steven and Van Durme, Benjamin", editor = "Gkatzia, Dimitra and Seddah, Djam{\'e}", booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations", month = apr, year = "2021", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.eacl-demos.19", doi = "10.18653/v1/2021.eacl-demos.19", pages = "149--159", abstract = "We present LOME, a system for performing multilingual information extraction. Given a text document as input, our core system identifies spans of textual entity and event mentions with a FrameNet (Baker et al., 1998) parser. It subsequently performs coreference resolution, fine-grained entity typing, and temporal relation prediction between events. By doing so, the system constructs an event and entity focused knowledge graph. We can further apply third-party modules for other types of annotation, like relation extraction. Our (multilingual) first-party modules either outperform or are competitive with the (monolingual) state-of-the-art. We achieve this through the use of multilingual encoders like XLM-R (Conneau et al., 2020) and leveraging multilingual training data. LOME is available as a Docker container on Docker Hub. In addition, a lightweight version of the system is accessible as a web demo.", } - R. Yasarla, Hamid Reza Vaezi Joze, and Vishal M. Patel, “Network Architecture Search for Face Enhancement,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{234681236, title = {Network Architecture Search for Face Enhancement}, author = {{R. Yasarla} and {Hamid Reza Vaezi Joze} and {Vishal M. Patel}}, year = 2021, month = {5}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/02c7dcedee24ae9ca55a96180fae7b7000009ad0}, } - Adam Kortylewski, Ju He, Qing Liu, Christian Cosgrove, Chenglin Yang, and A. Yuille, “Compositional Generative Networks and Robustness to Perceptible Image Changes,” in Annual Conference on Information Sciences and Systems, 2021.
[BibTeX] [Link]@inproceedings{233333658, title = {Compositional Generative Networks and Robustness to Perceptible Image Changes}, author = {{Adam Kortylewski} and {Ju He} and {Qing Liu} and {Christian Cosgrove} and {Chenglin Yang} and {A. Yuille}}, year = 2021, month = {3}, booktitle = {Annual Conference on Information Sciences and Systems}, url = {https://www.semanticscholar.org/paper/90acfd3862d20103cd925d470ca7f6bf83e3514f}, } - L. Moro-Velázquez, J. Gómez-García, N. Dehak, and J. Godino-Llorente, “New tools for the differential evaluation of Parkinson’s disease using voice and speech processing,” in IberSPEECH Conference, 2021.
[BibTeX] [Link]@inproceedings{232285765, title = {New tools for the differential evaluation of Parkinson's disease using voice and speech processing}, author = {{L. Moro-Velázquez} and {J. Gómez-García} and {N. Dehak} and {J. Godino-Llorente}}, year = 2021, month = {3}, booktitle = {IberSPEECH Conference}, url = {https://www.semanticscholar.org/paper/c76e00b4e7c3fa5774cb61a194535086f53b7802}, } - Yingda Xia, Dong Yang, Wenqi Li, A. Myronenko, Daguang Xu, Hirofumi Obinata, Hitoshi Mori, P. An, S. Harmon, E. Turkbey, B. Turkbey, B. Wood, F. Patella, Elvira Stellato, G. Carrafiello, A. Ierardi, A. Yuille, and H. Roth, “Auto-FedAvg: Learnable Federated Averaging for Multi-Institutional Medical Image Segmentation,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{233324290, title = {Auto-FedAvg: Learnable Federated Averaging for Multi-Institutional Medical Image Segmentation}, author = {{Yingda Xia} and {Dong Yang} and {Wenqi Li} and {A. Myronenko} and {Daguang Xu} and {Hirofumi Obinata} and {Hitoshi Mori} and {P. An} and {S. Harmon} and {E. Turkbey} and {B. Turkbey} and {B. Wood} and {F. Patella} and {Elvira Stellato} and {G. Carrafiello} and {A. Ierardi} and {A. Yuille} and {H. Roth}}, year = 2021, month = {4}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/7ba0e1864f094da06410af52166dc6c4e7a74adf}, } - Xing Di, He Zhang, and Vishal M. Patel, “Thermal-to-Visible Face Synthesis and Recognition,” in Advances in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{238912924, title = {Thermal-to-Visible Face Synthesis and Recognition}, author = {{Xing Di} and {He Zhang} and {Vishal M. Patel}}, year = 2021, booktitle = {Advances in Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/c02d3e2e51d8fcea46e36822b110b26c140d5d24}, } - Nataniel Ruiz, Adam Kortylewski, Weichao Qiu, Cihang Xie, Sarah Adel Bargal, A. Yuille, and S. Sclaroff, “Simulated Adversarial Testing of Face Recognition Models,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{235367664, title = {Simulated Adversarial Testing of Face Recognition Models}, author = {{Nataniel Ruiz} and {Adam Kortylewski} and {Weichao Qiu} and {Cihang Xie} and {Sarah Adel Bargal} and {A. Yuille} and {S. Sclaroff}}, year = 2021, month = {6}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/1f5ec5c5bc69ee850d31d70281322e8026c5bd52}, } - Nanxin Chen, Piotr Żelasko, L. Moro-Velázquez, J. Villalba, and N. Dehak, “Align-Denoise: Single-Pass Non-Autoregressive Speech Recognition,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{237634474, title = {Align-Denoise: Single-Pass Non-Autoregressive Speech Recognition}, author = {{Nanxin Chen} and {Piotr Żelasko} and {L. Moro-Velázquez} and {J. Villalba} and {N. Dehak}}, year = 2021, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/2161383af6d420450f69ada26f2e310e554750f8}, } - Piotr Żelasko, R. Pappagari, and N. Dehak, “What Helps Transformers Recognize Conversational Structure? Importance of Context, Punctuation, and Labels in Dialog Act Recognition,” in Transactions of the Association for Computational Linguistics, 2021.
[BibTeX] [Link]@inproceedings{235742745, title = {What Helps Transformers Recognize Conversational Structure? Importance of Context, Punctuation, and Labels in Dialog Act Recognition}, author = {{Piotr Żelasko} and {R. Pappagari} and {N. Dehak}}, year = 2021, month = {7}, booktitle = {Transactions of the Association for Computational Linguistics}, url = {https://www.semanticscholar.org/paper/f3173cd86ae95a53f44f0d1093e85df4988a459a}, } - Qing Liu, Vignesh Ramanathan, D. Mahajan, A. Yuille, and Zhenheng Yang, “Weakly Supervised Instance Segmentation for Videos with Temporal Mask Consistency,” in Computer Vision and Pattern Recognition, 2021.
[BibTeX] [Link]@inproceedings{232335658, title = {Weakly Supervised Instance Segmentation for Videos with Temporal Mask Consistency}, author = {{Qing Liu} and {Vignesh Ramanathan} and {D. Mahajan} and {A. Yuille} and {Zhenheng Yang}}, year = 2021, month = {3}, booktitle = {Computer Vision and Pattern Recognition}, url = {https://www.semanticscholar.org/paper/5b62f55ee46245b0c4e710efac9c7578666c68a3}, } - N. Higgins, Ambar Monjaras, Breanne D Yerkes, David F. Little, J. Nave-Blodgett, Mounya Elhilali, and J. Snyder, “Resetting of Auditory and Visual Segregation Occurs After Transient Stimuli of the Same Modality,” in Frontiers in Psychology, 2021.
[BibTeX] [Link]@inproceedings{236230536, title = {Resetting of Auditory and Visual Segregation Occurs After Transient Stimuli of the Same Modality}, author = {{N. Higgins} and {Ambar Monjaras} and {Breanne D Yerkes} and {David F. Little} and {J. Nave-Blodgett} and {Mounya Elhilali} and {J. Snyder}}, year = 2021, month = {6}, booktitle = {Frontiers in Psychology}, url = {https://www.semanticscholar.org/paper/5720f41225f3203b71d2b0d40a0776d9e7d5db15}, } - Xing Di, Shuowen Hu, and Vishal M. Patel, “Heterogeneous Face Frontalization via Domain Agnostic Learning,” in IEEE International Conference on Automatic Face & Gesture Recognition, 2021.
[BibTeX] [Link]@inproceedings{236087940, title = {Heterogeneous Face Frontalization via Domain Agnostic Learning}, author = {{Xing Di} and {Shuowen Hu} and {Vishal M. Patel}}, year = 2021, month = {7}, booktitle = {IEEE International Conference on Automatic Face & Gesture Recognition}, url = {https://www.semanticscholar.org/paper/cb854b48d871265a457a2fbb97d01ca40bb9f4c3}, } - Piotr Żelasko, Sonal Joshi, Yiwen Shao, J. Villalba, J. Trmal, N. Dehak, and S. Khudanpur, “Adversarial Attacks and Defenses for Speech Recognition Systems,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{232427815, title = {Adversarial Attacks and Defenses for Speech Recognition Systems}, author = {{Piotr Żelasko} and {Sonal Joshi} and {Yiwen Shao} and {J. Villalba} and {J. Trmal} and {N. Dehak} and {S. Khudanpur}}, year = 2021, month = {3}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/9d15685433a067c5beca67e5f6cc612b3dc29f66}, } - N. Mathioudakis, Mohammed S. Abusamaan, Ahmed F. Shakarchi, Sam Sokolinsky, Shamil Fayzullin, J. McGready, Mihail Zilbermint, S. Saria, and S. Golden, “Development and Validation of a Machine Learning Model to Predict Near-Term Risk of Iatrogenic Hypoglycemia in Hospitalized Patients,” in JAMA Network Open, 2021.
[BibTeX] [Link]@inproceedings{231194796, title = {Development and Validation of a Machine Learning Model to Predict Near-Term Risk of Iatrogenic Hypoglycemia in Hospitalized Patients}, author = {{N. Mathioudakis} and {Mohammed S. Abusamaan} and {Ahmed F. Shakarchi} and {Sam Sokolinsky} and {Shamil Fayzullin} and {J. McGready} and {Mihail Zilbermint} and {S. Saria} and {S. Golden}}, year = 2021, month = {1}, booktitle = {JAMA Network Open}, url = {https://www.semanticscholar.org/paper/c172a8a4adad62a49167e9a76e6a4951565a4894}, } - Jiteng Mu, Weichao Qiu, Adam Kortylewski, A. Yuille, N. Vasconcelos, and Xiaolong Wang, “A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation,” in IEEE International Conference on Computer Vision, 2021.
[BibTeX] [Link]@inproceedings{233240742, title = {A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representation}, author = {{Jiteng Mu} and {Weichao Qiu} and {Adam Kortylewski} and {A. Yuille} and {N. Vasconcelos} and {Xiaolong Wang}}, year = 2021, month = {4}, booktitle = {IEEE International Conference on Computer Vision}, url = {https://www.semanticscholar.org/paper/289f55883db1b91ff1c8d9e4a36dbdd6c3e2782e}, } - Prakhar Kaushik, Alex Gain, Adam Kortylewski, and A. Yuille, “Understanding Catastrophic Forgetting and Remembering in Continual Learning with Optimal Relevance Mapping,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{232013967, title = {Understanding Catastrophic Forgetting and Remembering in Continual Learning with Optimal Relevance Mapping}, author = {{Prakhar Kaushik} and {Alex Gain} and {Adam Kortylewski} and {A. Yuille}}, year = 2021, month = {2}, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/be864a16ec597c76d1ab36453d01471723a37bac}, } - Enayat Ullah, Tung Mai, Anup B. Rao, Ryan A. Rossi, and R. Arora, “Machine Unlearning via Algorithmic Stability,” in Annual Conference Computational Learning Theory, 2021.
[BibTeX] [Link]@inproceedings{232068763, title = {Machine Unlearning via Algorithmic Stability}, author = {{Enayat Ullah} and {Tung Mai} and {Anup B. Rao} and {Ryan A. Rossi} and {R. Arora}}, year = 2021, month = {2}, booktitle = {Annual Conference Computational Learning Theory}, url = {https://www.semanticscholar.org/paper/0fa360d5bb8ce649155c6816fd19e5bffac4e07c}, } - R. Pappagari, Jaejin Cho, Sonal Joshi, L. Moro-Velázquez, Piotr Żelasko, J. Villalba, and N. Dehak, “Automatic Detection and Assessment of Alzheimer Disease Using Speech and Language Technologies in Low-Resource Scenarios,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{239653935, title = {Automatic Detection and Assessment of Alzheimer Disease Using Speech and Language Technologies in Low-Resource Scenarios}, author = {{R. Pappagari} and {Jaejin Cho} and {Sonal Joshi} and {L. Moro-Velázquez} and {Piotr Żelasko} and {J. Villalba} and {N. Dehak}}, year = 2021, month = {8}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/7e3deabd44eccb0fe2823d8cecf1e182efeeb0f6}, } - N. Finkelstein, R. Adams, S. Saria, and I. Shpitser, “Partial Identifiability in Discrete Data With Measurement Error Supplementary Material.” 2021.
[BibTeX] [Link]@inproceedings{260439828, title = {Partial Identifiability in Discrete Data With Measurement Error Supplementary Material}, author = {{N. Finkelstein} and {R. Adams} and {S. Saria} and {I. Shpitser}}, year = 2021, booktitle = {}, url = {https://www.semanticscholar.org/paper/35bbd389372786ef73c102ba7fab0126ea4c1184}, } - Sonal Joshi, J. Villalba, Piotr Żelasko, Laureano Moro-Vel’azquez, and N. Dehak, “Adversarial Attacks and Defenses for Speaker Identification Systems,” in arXiv.org, 2021.
[BibTeX] [Link]@inproceedings{231693283, title = {Adversarial Attacks and Defenses for Speaker Identification Systems}, author = {{Sonal Joshi} and {J. Villalba} and {Piotr Żelasko} and {Laureano Moro-Vel'azquez} and {N. Dehak}}, year = 2021, booktitle = {arXiv.org}, url = {https://www.semanticscholar.org/paper/b595a080a4376bab6edd2e8b8c4bfa3cede54f3b}, } - Prithviraj Dhar, Joshua Gleason, A. Roy, C. Castillo, and R. Chellappa, “Supplementary Material-PASS.” 2021.
[BibTeX] [Link]@inproceedings{244312236, title = {Supplementary Material-PASS}, author = {{Prithviraj Dhar} and {Joshua Gleason} and {A. Roy} and {C. Castillo} and {R. Chellappa}}, year = 2021, booktitle = {}, url = {https://www.semanticscholar.org/paper/ee50fa46cd195e4b59330297d4285877906583b5}, } - J. Villalba, Sonal Joshi, Piotr Żelasko, and N. Dehak, “Representation Learning to Classify and Detect Adversarial Attacks Against Speaker and Speech Recognition Systems,” in Interspeech, 2021.
[BibTeX] [Link]@inproceedings{235790692, title = {Representation Learning to Classify and Detect Adversarial Attacks Against Speaker and Speech Recognition Systems}, author = {{J. Villalba} and {Sonal Joshi} and {Piotr Żelasko} and {N. Dehak}}, year = 2021, month = {7}, booktitle = {Interspeech}, url = {https://www.semanticscholar.org/paper/8abbc820db608654c4ba10203245c191566e7286}, } - Sarel van Vuuren and H. Hermansky, “SPEECH VARIABILITY IN THE MODULATION SPECTRAL DOMAIN.” 2021.
[BibTeX] [Link]