AI Center Seminar - AI for Robotics series - Prof. Qifeng Chen

The talk is jointly organized by the EPFL AI Center and the VITA Lab as part of the AI for Robotics seminar series.

Hosting professor: Prof. Alexandre Alahi (VITA)

Title
From Pixels to Physicality: Learning Motion Intelligence for Video Generation and Humanoid Interaction

Abstract
The evolution of generative AI has shifted from synthesizing images to modeling complex temporal dynamics in open domains. However, it remains challenging to generate and perform motion in a robust, physically feasible way in both virtual and physical worlds. In this talk, I will present a perspective on motion intelligence, the ability to generate and execute complex movements in diverse domains. I will demonstrate how scaling data and architectural priors enables us to build a generative bridge between digital content creation and embodied intelligence, ultimately moving towards AI that can seamlessly interact with people in the physical world.

Bio
Qifeng Chen is an Associate Professor at the Hong Kong University of Science and Technology. He received his PhD in Computer Science from Stanford University. His research interests include generative AI, computational photography, and embodied AI. He was named one of the "35 Innovators Under 35" and one of the top 100 AI young pioneers in China by MIT Technology Review. He received the Google Faculty Research Award. He served as an area chair for international conferences, including CVPR, NeurIPS, ICCV, ECCV, ACM Multimedia, and ICML. He is also serving as an associate editor of TPAMI.