Home
About
People
Publications
Seminars
- Seminar Calendar
- Seminar Videos
Workshops
HLT Masters
Resources

Graduate Students

A-J
K-Q
R-Z

Carlos Aguirre | Website | Publications
Niyati Bafna | Website | Publications
Elina Baral
Andrew Blair-Stanek | Website | Publications
Isabel Cachola | Website | Publications
Tianyu Cao | Website | Publications
Cameron Carpenter
Jonathan Chang
Kaavya Chaparal
Alexandra DeLucia | Website
Leo Du | Publications
Seth Ebner | Website
Anna Favaro
William Fleshman
Dongji Gao | Publications
Yaohan Guan
Sophia Hager
Keith Harrigian| Website | Publications
Amir Hussein
Heyuan Huang
Ruizhe Huang
Zili Huang
Maliha Jahan | Publications
Zhengping Jiang | Website | Publications
Sonal Sunil Joshi
William Jurayj

Aleem Khan | Website | Publications
Junhyeok Lee
Sabrina Li
Tianjian Li
Xinyuan (Henry) Li | Website | Publications
Brian Lu
Yen-Ju Lu
Marc Marone | Website | Publications
Alexander Martin
David Mueller | Website | Publications
Bismarck Odoom
Jiefu Ou
Candice Laine Penelton
Guanghui Qin | Website | Publications

Qinwan Rabbani
Krithika Ramesh | Website | Publications
Kathryn Ricci
Elizabeth Salesky | Website | Publications
Kate Sanders | Website | Publications
Yiwen Shao
Rafael Rivera Soto
Nate Robinson
Jamie Rosas-Smith
Kaiser Sun | Website | Publications
Steven Tan
Neha Verma
Andrew Wang
Helin Wang
Miriam Wanner
Nathaniel Weir
Orion Weller | Website | Publications
Rachel Wicks
Cihan Xiao
He Xiluo
Haoran Xu | Website | Publications
Jack Zhang
Xuan Zhang | Website | Publications
Mian Zhong

People

Faculty
- Affiliated Faculty
- Former Faculty
Postdocs
- Current Postdocs
- Former Postdocs
Graduate Students
Staff
Alumni

Subscribe to the CLSP Seminars Mailing List

Upcoming Seminars

There are no upcoming events.

Add

Add to Timely Calendar
Add to Google
Add to Outlook
Add to Apple Calendar
Add to other calendar
Export to XML

Seminar Videos

📍 Live from FIT, Brno University of Technology (Czech Republic), room E112
🕘 July 1st, 2025 — 11:00 CEST
🎙️ Barbara Schuppler (TU Graz, Austria)

In recent years, conversational speech has become a major focus in speech science and technology. As dialogue systems evolve from transactional tools into socially interactive agents, they demand increasingly accurate automatic speech recognition (ASR). At the same time, conversational data offers unique insights into human speech processing. Drawing on the cross-layer optimization principle from communications engineering, I adopt a similar view of how meaning is accessed across multiple levels of speech information. In this talk, I present findings from my group’s work on integrating pronunciation and prosodic variation into ASR for conversational speech. Our hybrid approach—combining data-driven and knowledge-based methods—proves especially effective in low-resource settings. While transformer-based models often outperform classical systems, the latter still excel with short, fragmented utterances when paired with linguistic knowledge. Beyond ASR, our methods inform fields like pathological speech analysis, dementia prediction, and assistive speech technologies.

Bio: Barbara Schuppler studied Physics and Spanish Philology at the University of Graz and the Universidad Autónoma de Madrid, completing a diploma thesis in experimental physics in 2007. She conducted her dissertation within the Marie-Curie RTN "Sound-to-Sense" at Radboud University Nijmegen, with research visits at NTNU Trondheim. After working as teacher at the Graz International Bilingual School, she was awarded an FWF Hertha-Firnberg Grant in 2012 and joined the Signal Processing and Speech Communication Laboratory at TU Graz. Now Associate Professor at TU Graz, her research interests include the investigation of methods for quantitative analyses of prosody and pronunciation variation in conversational speech, the integration of gained phonetic and linguistic knowledge into speech technology, with a specific focus on applications in the educational and healthcare sector.

[camera] JSALT 2025 - Plenary - Barbara Schuppler: Cross-layer Models for Low-Res Conversational ASR

📍 Live from FIT, Brno University of Technology (Czech Republic), room E112
🕘 July 1st, 2025 — 11:00 CEST
🎙️ Barbara Schuppler (TU Graz, Austria)

In recent years, conversational speech has become a major focus in speech science and technology. As dialogue systems evolve from transactional tools into socially interactive agents, they demand increasingly accurate automatic speech recognition (ASR). At the same time, conversational data offers unique insights into human speech processing. Drawing on the cross-layer optimization principle from communications engineering, I adopt a similar view of how meaning is accessed across multiple levels of speech information. In this talk, I present findings from my group’s work on integrating pronunciation and prosodic variation into ASR for conversational speech. Our hybrid approach—combining data-driven and knowledge-based methods—proves especially effective in low-resource settings. While transformer-based models often outperform classical systems, the latter still excel with short, fragmented utterances when paired with linguistic knowledge. Beyond ASR, our methods inform fields like pathological speech analysis, dementia prediction, and assistive speech technologies.

Bio: Barbara Schuppler studied Physics and Spanish Philology at the University of Graz and the Universidad Autónoma de Madrid, completing a diploma thesis in experimental physics in 2007. She conducted her dissertation within the Marie-Curie RTN "Sound-to-Sense" at Radboud University Nijmegen, with research visits at NTNU Trondheim. After working as teacher at the Graz International Bilingual School, she was awarded an FWF Hertha-Firnberg Grant in 2012 and joined the Signal Processing and Speech Communication Laboratory at TU Graz. Now Associate Professor at TU Graz, her research interests include the investigation of methods for quantitative analyses of prosody and pronunciation variation in conversational speech, the integration of gained phonetic and linguistic knowledge into speech technology, with a specific focus on applications in the educational and healthcare sector.

[slides] JSALT 2025 - Plenary - Barbara Schuppler: Cross-layer Models for Low-Res Conversational ASR

Humans can listen to a desired sound within a complex acoustic scene consisting of a mixture of various sounds. This phenomenon, called the cocktail party effect or selective hearing, enables us to listen to an interlocutor in a noisy cafe, focus on a particular instrument in a song, or notice a siren on the road.

In this talk, I will discuss target speech/sound extraction (TSE), which isolates the speech signal of a target speaker or a target sound from a mixture of several speakers or sounds using clues that identify the target in the mixture. Such clues might be a spatial clue indicating the direction of the target, a video of the target, or a prerecorded enrollment audio from which the speaker’s voice or the target sound characteristics can be derived. I will introduce the foundation and present recent research on neural-based TSE for speech and arbitrary sounds.

Bio: Marc Delcroix is a Distinguished Researcher with NTT Communication Science Laboratories, NTT Corporation, Japan. He received the M.Eng. degree from the Free University of Brussels, Brussels, Belgium, and Ecole Centrale Paris, Paris, France, in 2003, and the Ph.D. degree from the Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Japan, in 2007. His research interests include various aspects of speech and audio processing, such as target speech and sound extraction, speech enhancement, robust speech recognition, model adaptation, and speaker diarization.

He is a member of the CHiME challenge steering committee, AASP-TC, a past member of the SL-TC from 2018 to 2023 and the organizing committees of the REVERB Challenge 2014, the ASRU 2017, and SLT 2022.

[camera] JSALT 2025 - Plenary Talk - Marc Delcroix: Neural Target Speech and Sound Extraction

Unlocking Audio AI: Building a Massive Open Dataset for Instruction-Tuned Audio-Text Foundation Models

Recent breakthroughs in language models have shown us the remarkable capabilities unlocked by large-scale instruction tuning—models trained to deeply understand and precisely respond to human instructions. Can we achieve the same revolution in the audio domain?

This talk will outline a vision and practical roadmap for collaboratively building a massive, open, and richly annotated audio-text dataset. Starting from millions of permissively licensed snippets of speech, diverse music genres, and environmental sounds, we'll create detailed captions describing each sound both acoustically and contextually. Using advanced generative AI such as Gemini, we'll orchestrate these individual samples into billions of carefully annotated soundscapes with accurate, timestamped event labels.

Such a dataset will enable transformative progress across multiple tasks, including audio-to-text transcription, nuanced sound-event detection, audio generation, and instruction-tuned multimodal voice assistants. Openness and permissive licensing will ensure the resulting dataset and models remain accessible, reproducible, and extensible by the global research community.

*Biosketch*: Christoph Schuhmann is the co‑founder and Organizational Lead of LAION (Large‑Scale Artificial Intelligence Open Network), a German non‑profit committed to advancing open‑source development in machine learning and AI. Under his volunteer leadership, LAION has released landmark datasets LAION‑400M (2021) and the 5‑billion‑pair LAION‑5B (2022)—which underpin influential open‑source models like Stable Diffusion He holds a master’s in physics and computer science and works as a high‑school teacher in Germany, specializing in both subjects. Christoph has a strong background in reform‑oriented education, having previously produced the “Schools of Trust” documentary and supported initiatives in alternative schooling

In research, he leads LAION’s efforts in open multimedia learning. His team co‑authored the NeurIPS 2022 paper introducing the LAION‑5B dataset, and he contributed to several recent works, including a 2025 paper on freeing scientific knowledge from copyright via LLMs and he also led the developement of EmoNet, an open-source suite of tools and datasets (EmoNet‑Face, EmoNet‑Voice, Empathic Insight models, BUD‑E Whisper) designed to enable emotionally intelligent AI capable of nuanced facial and vocal affect analysis

Christoph is a staunch advocate for open science in AI. He co‑initiated OpenAssistant—now one of the most significant community‑driven open‑source chatbot projects—and actively promotes transparent, decentralized development of foundational AI technologies.

[camera] JSALT 2025 - Seminar with Christoph Schuhmann (LAION)

🎉 JSALT 2025 – Opening Day (June 23rd)
📍 Live from FIT, Brno University of Technology (Czech Republic)
🕘 All times below are expected video timestamps

Kick off JSALT 2025 with the opening remarks and team presentations from this year’s cutting-edge research workshop on speech and language technology. Watch live from 9 A.M. (CEST) or catch up anytime using the timestamps below!

⏱️ Timestamps
00:10 — Call to Order – Honza Černocký
00:15 — Welcome Remarks – BUT Rector & FIT Representative
00:20 — How We Got Here: A History of the Summer Workshops – Sanjeev Khudanpur
00:40 — EMMA: End-to-end Multi-channel Multi-talker ASR – Lukáš Burget
01:05 — Discussion & Q&A
01:10 — Play Your Part: LLM Role-playing Agents That Stick to Their Role – Ricard Marxer
01:35 — Discussion & Q&A
01:40 — Break
01:55 — AuGI: Towards Audio General Intelligence – Ramani Duraiswami
02:20 — Discussion & Q&A
02:25 — TTS4ALL: TTS in Low Resource Scenarios – Yannick Estève
02:50 — Discussion & Q&A
02:55 — JSALT 2025 Organizational Information – Honza Černocký

🔴 Streamed live from Room E112, FIT BUT (Bozetechova 2, Brno)

#JSALT2025 #SpeechTech #ASR #TTS #LLM #ArtificialIntelligence #LiveFromBrno

[slides] JSALT Summer Workshop 2025 - Opening Day (June 23rd)

Load More... Subscribe

Johns Hopkins University, Whiting School of Engineering

Center for Language and Speech Processing
Hackerman 226
3400 North Charles Street, Baltimore, MD 21218-2680

(410) 516-4237
[email protected]

Whiting School of Engineering
Johns Hopkins University

Privacy Statement
Legal

© 2025 Johns Hopkins University

Center for Language and Speech Processing