About Me: (Link to CV) I am a Ph.D. student in the department of Electrical and Computer Engineering. I received my bachelor’s degree from Department of Electronic Engineering, Tsinghua University. I am working with Dr. Sanjeev Khudanpur and Dr. Daniel Povey on keyword spotting and speech recognition.
Research Interests: Large Vocabulary Conversational Speech Recognition, Keyword Search (on speech corpus, also known as keyword spotting, wake-up words, etc.), Deep Neural Networks, Natural Language Processing, Machine Learning.Generally I’m interested in techniques that can improvement speech recognition and keyword search performance. But I’m also very interested in coding related fancy projects.
Guoguo Chen, Hainan Xu, Minhua Wu, Daniel Povey and Sanjeev Khudanpur. “Pronunciation and silence probability modeling for ASR,” Interspeech 2015. [pdf]
Vijayaditya Peddinti, Guoguo Chen, Daniel Povey and Sanjeev Khudanpur. “Reverberation robust acoustic modeling using i-vectors with time delay neural networks,” Interspeech 2015. [pdf]
Hainan Xu, Guoguo Chen, Daniel Povey and Sanjeev Khudanpur. “Modeling phonetic context with non-random forests for speech recognition,” Interspeech 2015. [pdf]
Guoguo Chen, Carolina Parada, Tara N. Sainath. “Query-by-example keyword spotting using long short-term memory networks,” ICASSP 2015. [pdf]
Vassil Panayotov, Guoguo Chen, Daniel Povey and Sanjeev Khudanpur. “LibriSpeech: an ASR corpus based on public domain audio books,” ICASSP 2015. [pdf]
Jan Trmal, Guoguo Chen, et al. “A keyword search system using open source software,” SLT 2014. [pdf]
Chunxi Liu, Aren Jansen, Guoguo Chen, Keith Kintzley, Jan Trmal and Sanjeev Khudanpur. “Low-resource open vocabulary keyword search using point process models,” Interspeech 2014. [pdf]
Justin Chiu, Yun Wang, Jan Trmal, Daniel Povey, Guoguo Chen and Alexander Rudnicky. “Combination of FST and CN search in spoken term detection,” Interspeech 2014. [pdf]
Guoguo Chen, Carolina Parada, Georg Heigold. “Small-footprint keyword spotting using deep neural networks,” ICASSP 2014. [pdf]
Guoguo Chen, Oguz Yilmaz, Jan Trmal, Daniel Povey, Sanjeev Khudanpur. “Using proxies for OOV keywords in the keyword search task,” ASRU 2013. [pdf]
Guoguo Chen, Sanjeev Khudanpur, Daniel Povey, Jan Trmal, David Yarowsky and Oguz Yilmaz. “Quantifying the value of pronunciation lexicons for keyword search in low resource languages,” ICASSP 2013. [pdf]
Reviewer of EURASIP Journal on Audio, Speech and Music Processing
Google Speech Research Group, MTV, Summer 2013: Worked with Carolina Parada, Georg Heigold, Alex Gruenstein
Developed a novel keyword spotting (wake-up word) framework based on deep neural network. The proposed framework has a small memory footprint, low computational cost, and high precision, and is appropriate for devices such as smartphones and tablets.
Google Speech Research Group, MTV, Summer 2014: Worked with Carolina Parada, Tara Sainath, Rohit Prabhavalkar
Developed a novel long short-term memory (LSTM) recurrent neural network based feature extractor for query-by-example (QbyE) keyword spotting (wake-up word). This keyword spotting can let users to specify their own keywords and has small memory footprint, low computational cost as well as high precision. It is suitable for devices such as smartphones and tablets.
Key Phrase Detection. GP-19639-00-US (16113-5295001).
Learning for Deep Neural Networks. GP019885-00-PR (16113-5314P01)
User Specified Keyword Spotting Using Long Short Term Memory Neural Network Feature Extractor. GP-22850-00-US (16113-6564001)
It is my hornor to work with great researchers along my Ph.D. life. Some of them are:
Sanjeev Khudanpur, advisor, Johns Hopkins University
Daniel Povey, co-advisor, Johns Hopkins University
Jan Trmal, Radical Team, Johns Hopkins University
Aren Jansen, Radical Team, Johns Hopkins University
David Yarowsky, Radical Team, Johns Hopkins University
Carolina Parada, Google Inc.
Georg Heigold, Google Inc.
Alex Gruenstein, Google Inc.
Tara Sainath, Google Inc.
Rohit Prabhavalkar, Google Inc.
I’ve implemented Dogan Can and Murat Saraclar’s lattice indexing paper for keyword search. Here is the download link (slightly out-of-date, for the latest version please refer to Kaldi’s keyword search module). Please let me know if you have any problem using it. Simple examples will be available upon request.
I’m a big fan of the Baltimore Ravens! Ups or downs, I’m always a raven 🙂
I participated in the 2006 Beijing international Marathon and finished the 42.195km within 4 hour and a half! I’m preparing one here in Baltimore.
The Center for Language and Speech Processing
The Johns Hopkins University
Hackerman Hall 322
3400 North Charles Street
Baltimore, MD 21218
* Telephone: (410) 516-4237 * Fax: (410) 516-5050 * E-mail: guoguo at jhu.edu