Xiaoshuai "Jet" Zhang

Xiaoshuai Zhang is a Research Scientist at sudoAI, where he is a founding member and research lead of the foundation model group. His current focus is on scaling up the embodied AI data to support faster and more efficient VLA and world model training. He received his Ph.D. in CS from UC San Diego in 2024 (w/ Prof. Hao Su) and a B.S. in CS from Peking University in 2019. He worked / interned at Apple, Adobe, Google, and Hillbot, and was a visiting scholar at Stanford (w/ Prof. Leonidas Guibas). Xiaoshuai’s research spans embodied AI models, sim2real, 3D vision, foundation models, and graphics.

For publications, visit Google Scholar. Feel free to connect if you are interested in collaborating.

Area Chair: CVPR'25 '26
Reviewer: CVPR, ICCV, ECCV, ICLR, ICML, ICRA, IROS, ACM MM, NeurIPS, SIGGRAPH Asia, TPAMI, IJCV, TVCG, RA-L, T-RO