Published numerous papers such as 'Autoregressive Distillation of Diffusion Transformers', 'FlexiDiT: Your Diffusion Transformer Can Easily Generate High-Quality Samples with Less Compute', etc.; Champion of ICCV 2021 UVO Open-World Segmentation Challenge (both Image-/Video-track); Champion of ECCV 2018 COCO Keypoint Detection Challenge.
Research Experience
Serves as a Research Scientist at Meta; involved in multiple research projects, including image generation, human motion prediction, object segmentation, etc.
Education
Ph.D. from ENPC, supervised by Prof. Vincent Lepetit; Master/Bachelor from Beihang University; Double degree from ECLyon.
Background
Research Scientist at Meta - Superintelligence Labs; Research interests include computer vision and machine learning.
Miscellany
Academic service includes reviewing for journals/conferences such as TPAMI, IJCV, CVPR, ECCV, ISMAR, WACV, BMVC.