Prof. WANG, Liming 汪 黎 明 教 授
Assistant Professor
PhD, BE (University of Illinois Urbana-Champaign)
Office: Room 601F, Ho Sin Hang Engineering Building
Tel: (852) 3943-9618
Email: lmwang@se.cuhk.edu.hk
Biography
Liming Wang is an assistant professor in the Department of Systems Engineering and Engineering Management at The Chinese University of Hong Kong. Before joining CUHK, he was a postdoctoral researcher at MIT CSAIL. He received his Ph.D. in Electrical and Computer Engineering from the University of Illinois Urbana–Champaign.
His research focuses on self-supervised and multimodal learning, with applications in speech and language technology, generative modeling, and healthcare. He develops algorithms and theoretical frameworks for learning from weakly paired or unpaired multimodal data, with a particular emphasis on alignment, decipherment, and representation learning.
His work has advanced low-resource speech-to-text, text-to-speech, and speech-to-sign language translation by uncovering latent correspondences across speech, text, images, and sign language videos. More broadly, his research seeks to build AI systems that can understand, generate, and connect human communication across modalities and languages. His work has appeared in leading venues including NeurIPS, ACL, Interspeech, ICASSP, and IEEE/ACM TASLP.
He is the recipient of the Robert T. Chien Memorial Award in Electrical Engineering and the Rambus Computer Engineering Fellowship, and he has served as a reviewer for Interspeech and IEEE/ACM TASLP. His current research focuses on developing principled approaches for understanding and utilizing modern generative models for speech, with the goal of addressing fundamental challenges in alignment, decipherment, and representation learning.
Research Interests
- Self-supervised and unsupervised speech representation learning
- Multimodal learning across speech, text, vision, audio, and unpaired modalities
- Generative models for speech, audio, language, and multimodal data
- Health and clinical applications of speech, audio, and multimodal AI
- Theory-informed machine learning, including representation learning, self-supervised learning and generative modeling
Selected Publications
- Liming Wang, Muhammad Jehanzeb Mirza, Yishu Gong, Yuan Gong, Jiaqi Zhang, Brian H. Tracey, Katerina Placek, Marco Vilela, James Glass. Can Diffusion Models Disentangle? A Theoretical Perspective. Neural Information Processing Systems (NeurIPS), 2025.
- Liming Wang, Mark Hasegawa-Johnson, Chang D. Yoo. Unsupervised Speech Recognition with N-skipgram and Positional Unigram Matching. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2024.
- Liming Wang, Mark Hasegawa-Johnson and Chang Yoo. A Theory of Unsupervised Speech Recognition. Annual Meeting of the Association for Computational Linguistics (ACL), 2023.
- Junrui Ni*, Liming Wang*, Heting Gao*, Kaizhi Qian, Yang Zhang, Shiyu Chang and Mark Hasegawa-Johnson. Unsupervised Text-to-Speech Synthesis by Unsupervised Automatic Speech Recognition. Conference of the International Speech Communication Association (Interspeech), 2022. *equal contribution.
- Liming Wang, SiyuanFeng, Mark Hasegawa-Johnson and Chang Yoo. Self-supervised Semantic-driven Phoneme Discovery for Zero-resource Speech Recognition. Annual Meeting of the Association for Computational Linguistics (ACL), 2022.
- Liming Wang, Mark Hasegawa-Johnson. A DNN-HMM-DNN Hybrid Model for Discovering Word-like Units from Spoken Captions and Image Regions. Conference of the International Speech Communication Association (Interspeech), 2020.