Tianhai Liang

Harbin Institute of Technology, Tsinghua University

Papers

Total Citations

H-Index

About

Tianhai Liang is a robotics and computer vision researcher whose work bridges the gap between perception and dexterous manipulation. His key research areas include 3D garment modeling, robotic catching, and mobile manipulation. Liang’s most notable contribution is **ClothesNet**, a large-scale, information-rich 3D garment model repository containing around 4,400 models across 11 categories, annotated with features, boundary lines, and keypoints. This dataset, cited over 13 times, provides a foundational resource for computer vision and robotic cloth-handling tasks, enabling advances in garment simulation, recognition, and manipulation. In a striking demonstration of dynamic control, Liang also led the development of **Catch It!**, a system that enables mobile dexterous hands to catch thrown objects in flight—a challenging task requiring agile motion, large spatial awareness, and diverse object interaction. This work, published in 2025, has already garnered 6 citations and showcases his ability to tackle real-world, time-critical robotic skills. Through these contributions, Liang is shaping how robots perceive and physically interact with complex, deformable objects and dynamic environments.