Tingwu Wang

Vector Institute, University of Toronto, Nvidia (United Kingdom)

Papers

Total Citations

112

H-Index

About

Tingwu Wang is a researcher working at the intersection of physics-based simulation, computer graphics, and robotics, with a focus on developing intelligent systems that can learn realistic motion and adaptive morphologies. His most influential contribution, "Physics-based Human Motion Estimation and Synthesis from Videos," has garnered 77 citations and represents a significant breakthrough in reducing the dependency on costly motion capture data. By proposing a framework for training generative models of physically plausible human motion directly from video, Wang opened new avenues for accessible and scalable motion synthesis applicable to gaming, graphics, and robotic simulation environments. His work on "Neural Graph Evolution: Towards Efficient Automatic Robot Design" (33 citations) tackles the longstanding challenge of automated robot morphology discovery, addressing the combinatorial complexity of design search spaces through neural graph-based evolutionary strategies. Together, these contributions reflect Wang's broader mission to make sophisticated physical intelligence more attainable — whether by grounding human motion learning in video rather than lab-captured data, or by automating the engineering intuition traditionally required in robotics. His research holds meaningful implications for students and practitioners seeking to build more adaptive, data-efficient AI systems across simulation and embodied intelligence domains.