Published multiple papers, including but not limited to: 'HERMES: Human-to-Robot Embodied Learning from Multi-Source Motion Data for Mobile Dexterous Manipulation', 'Learning to Manipulate Anywhere: A Visual Generalizable Framework For Reinforcement Learning'. Among them, 'Pre-Trained Image Encoder for Generalizable Visual Reinforcement Learning' was accepted by NeurIPS 2022 and highlighted as a Spotlight paper.
Research Experience
Conducting PhD research at IIIS, Tsinghua University; Served as a research intern at UCSD.
Education
PhD student at Institute for Interdisciplinary Information Sciences (IIIS), Tsinghua University, advised by Prof. Huazhe Xu; Research intern at UCSD, advised by Prof. Xiaolong Wang; Master's degrees from Tsinghua University, Shenzhen Institute Graduate School, Big Data Engineering Track. Received Outstanding Thesis Award of Tsinghua University and Outstanding Graduate of Beijing.
Background
Broadly interested in research on generalization, representation learning, sim2real for RL agents, dexterous hands, robotics, and Embodied AI. The dream is to empower AI with the capacity to adapt and generalize in complex real-world environments.
Miscellany
Avid fan of the NBA and San Antonio Spurs, enjoys analyzing teams' labor contracts and each player's offensive and defensive statistics. Also loves singing and playing the piano.