Tianjun Xiao

Shanghai Artificial Intelligence Laboratory

Papers

Total Citations

H-Index

About

Tianjun Xiao is a leading researcher in computer vision and multimodal AI, with a focus on advancing video understanding, segmentation, and open-world perception. His most notable contribution is the development of VideoSAM, a pioneering framework that extends the Segment Anything Model (SAM) to open-world video segmentation—a critical capability for robotics and autonomous driving. By enabling continuous object association across video frames without requiring task-specific fine-tuning, Xiao’s work addresses a fundamental challenge in dynamic scene understanding. His research has garnered significant attention, with papers accumulating over 2,000 citations, reflecting its impact on both academic and applied domains. Xiao’s innovations have been recognized through top-tier conference publications and collaborations with leading tech labs, positioning him as a key figure in bridging static image segmentation and real-time video analysis. His work not only advances foundational AI but also directly supports practical systems in autonomous navigation and interactive robotics, making him a vital contributor to the next generation of intelligent visual agents.