Yang Fei费阳
World models and neural simulators for robotics.
I am an incoming Ph.D. student in Computer Science at the University of Toronto, advised by Prof. Sanja Fidler.
I study motion-centric video generation, with an emphasis on preserving structure and dynamics over time. My goal is to build interactive, physically grounded world models that help robots predict consequences and plan before acting.
Previously, I studied Computer Science and Mathematics at HKUST and was an exchange student at the University of Washington, working with Prof. Qifeng Chen and Prof. Ranjay Krishna.
News
- I will join the University of Toronto as a Computer Science Ph.D. student, advised by Prof. Sanja Fidler.
- Our new preprint, “Structure from Tracking” (first author), is now available. See the project page.
- VideoVAE+ (co-first author) has been accepted to ICCV 2025. Code and weights are available here.
Older news
- Received the The Hong Kong, China – Asia-Pacific Scholarship.
- Started an exchange program at the University of Washington (Seattle).
Selected Publications
All publicationsStructure From Tracking: Distilling Structure-Preserving Motion for Video Generation
Under Review
Distilling motion from a tracking model to preserve structure in generated videos.
Large Motion Video Autoencoding with Cross-modal Video VAE
ICCV 2025
Combining spatial and temporal compression with text guidance for high-fidelity video autoencoding.
Education
University of Toronto
Ph.D. in Computer Science · Advisor: Prof. Sanja Fidler
University of Washington
Exchange Student, College of Engineering · GPA: 4.0 / 4.0
Hong Kong University of Science and Technology
B.Sc. in Computer Science and Mathematics · GPA: 4.0 / 4.3 · Top 2 in CS & Math
Selected Honors
- Kerry Holdings Limited UROP Award
- HKSAR Government Scholarship
- The Hong Kong, China – Asia-Pacific Scholarship