Chieh-Hsin (Jesse) Lai (賴杰昕)
- Research Scientist/Tech Leader at Sony AI's Music Foundation Model Team;
- Visiting Assistant Professor at Department of Applied Mathematics, National Yang Ming Chiao Tung University, Taiwan;
- Ph.D. in Mathematics from University of Minnesota, Twin Cities
Currently
I am a research scientist and tech lead at Sony AI’s Music Foundation Model Team, focusing on deep generative modeling and robustness, particularly the foundations of diffusion models in collaboration with Prof. Stefano Ermon. Additionally, I am a Visiting Assistant Professor at National Yang Ming Chiao Tung University’s Applied Mathematics Department, where I collaborate and supervise students with Prof. Ming-Chih Lai on deep learning methods for partial differential equations.
My current research centers on enhancing diffusion models, emphasizing sample quality/diversity, sampling speed, and controllable generation. I am also interested about inverse problems, particularly in media restoration, utilizing generative models. Additionally, I explore mathematically explainable AI, aiming to unravel the intricacies of artificial intelligence through rigorous mathematical frameworks.
Previously
I earned my Ph.D. in Mathematics from the University of Minnesota – Twin Cities in May 2021. Under the supervision of Prof. Gilad Lerman and in close collaboration with Prof. Dongmian Zou, my research centered on constructing outlier robustness models and developing methodologies for outlier detection. I obtained my Bachelor’s degree in Mathematics from National Tsing Hua University, Taiwan in 2015. Here is my Mathematics Genealogy.
news
Dec 10, 2024 | 🔥 Excited to organize and present a tutorial on Diffusion Models @ ICASSP 2025! Details coming soon! |
---|---|
Nov 10, 2024 | 🔥 Organizing and giving tutorial on Diffusion Models @ ISMIR 2024! |
Oct 10, 2024 | 🥳 NeurIPS 2024 accepted papers: PaGoDA, GenWarp! |
Jan 17, 2024 | 🥳 ICLR 2024 accepted papers (3/3): CTM, MPGD, SAN! |
Oct 22, 2023 | 🔥 Sander’s great post summarizes perfectly our theoretical foresee which connects different consistency notions in Diffusion Models! |