Puhao Li
Puhao Li is a Ph.D. student in the Department of Automation at Tsinghua University, advised by Prof. Song-Chun Zhu, and a research intern at the Beijing Institute for General Artificial Intelligence (BIGAI). His research focuses on the intersection of robotic manipulation and 3D computer vision, with particular emphasis on dexterous grasping, bimanual manipulation, and 3D scene understanding. He is known for his work on ManipTrans, a method for transferring human bimanual manipulation skills to dexterous robotic hands, and the DexManipNet dataset, as well as contributions to generalizable dexterous grasping and agent-agnostic manipulation representations.
“TeleDexter trains a single RL controller that handles in-hand translation, rotation, finger gaiting, and long-horizon tool use”
Source→“TeleDexter achieves 75.2% average SR and 87.1% average TP across all seven tasks, while all baselines near-uniformly fail”
Source→“Despite this substantial morphological gap, LeapHand achieves 60.0–73.3% SR on the reorientation tasks, confirming that the framework generalizes across embodiments without re-collecting human reference motions”
Source→“Puhao Li, Tsinghua University / BIGAI, Lead author (equal contribution). Associated with the State Key Lab of General Artificial Intelligence at BIGAI”
Source→“State Key Lab of General Artificial Intelligence at BIGAI. Co-authored ManipTrans... a sustained research program on dexterous grasping and manipulation”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.