I am a Tenure-Track Assistant Professor at Institute for Interdisciplinary Information Sciences (IIIS), Tsinghua University. I am leading the Tsinghua Embodied AI Lab (TEA Lab, logo), where we build robots and then bring intelligence to robots.
I was a postdoctoral researcher at Stanford Vision and Learning Lab (SVL) advised by Prof. Jiajun Wu. I obtained my Ph.D. in Berkeley AI Research (BAIR) advised by Prof. Trevor Darrell. I obtained my bachelor degree from Tsinghua University (major in EE, minor in Management).
I am hiring self-motivated Postdocs, Ph.D. students, Undergrads, and Research Interns at Tsinghua (base: Beijing, Shanghai). See more details in Chinese here.
I am also hiring talented engineers and interns in an embodied AI startup.
Please email me with your CV directly.
I support Slow Science.
Research Topics
I am interested in Embodied AI: Reinforcement Learning, Robotics, and Computer Vision/Touch. Specifically, my research focuses on modeling the dynamics of the world, leveraging/finding human priors for policy learning, and further enabling algorithms to learn in a sample-efficient manner and generalize to unseen scenarios. I am also interested in solving complex real robot applications with deep learning and reinforcement learning.
News
Two papers (Make-An-Agent, KeyGrid) are accepted by NeurIPS'24.
I serve as an associate editor for ICRA'25.
Three papers (Maniwhere, RiEMann, GenSim2) are accepted by CoRL'24.
I serve as an area chair for ICLR'25.
I won Yunfan "Brilliant Star" Award in World Artificial Intelligence Conference (WAIC) 2024.
Two papers (RoboABC, Diffusion Reward) are accepted by ECCV'24.
V-Loco-Manip is accepted by IROS'24.
DP3 is accepted by RSS'24.
Six papers (Rethinking Transformers in Solving POMDPs, OMPO, Premier-TACO, BEE, OBAC, ACE) are accepted by ICML'24.
I serve as a meta-reviewer in Pioneers Workshop in RSS'24.
Three papers (Arraybot, DeformNet, HIRO Hand) are accepted by ICRA'24.
Six papers (DrM (spotlight), GenSim(spotlight), Uni-O4, LaMo, COPlanner, MorphMaze) are accepted by ICLR'24.
I serve as a reviewer in Science Robotics.
I serve as a Senior PC in IJCAI'24.
9DTact is accepted by RA-L with a presentation at ICRA'24.
I served as session chairs at DAI and X-AGI conference. I also will give a tutorial on RL at DAI.
I gave a talk about Embodied AI at YiXi@Shenzhen.
RoboCraft (3D version) is accepted by IJRR.
Our paper RoboCook won best system paper award at CoRL'23.
Our paper GenSim won outstanding paper award in LangRob workshop at CoRL'23.
I serve as an associate editor in ICRA'24.
Six papers (RL-ViGen, TACO, LfVoid, MoVie, H-InDex, CEIL) are accepted by NeurIPS'23 including 1 in dataset and benchmark track.
OTAS is accepted by WACV'24.
RoboCook is accepted by CoRL'23 (oral).
Pretrain-or-not-pretrain is accepted by ICML'23.
I will serve as an SPC member in IJCAI'23.
Four papers (DeFog, EAR, ScalingDiffPlanning, no-ensemble-mbrl) are accepted by ICLR'23.
Four papers(USEEK, DTact, EIL, Bimanual Handover) are accepted by ICRA'23.
Two papers are accepted by NeurIPS'22 (PIE-G, EMAPP, 2 spotlights).
One paper (see-hear-feel) is accepted by CoRL'22.
Our paper SUMMON is accepted by Siggraph Asia'22.
Our paper "Towards Learning to Play Piano with Dexterous Hands and Touch" is accepted by IROS'22.
Our RSS paper "RoboCraft" is covered by MIT Tech Review and HAI.
Two papers are accepted by ICML'22.
Our paper "Don't Touch What Matters" is accepted by IJCAI'22.
I gave a talk at UC San Diego.
Our paper RoboCraft is accepted by RSS'22.
One paper is accepted by ICLR'22 (spotlight).
Two papers are accepted by NeurIPS'21.
I became Dr. Xu on May 15, 2021.
I gave a talk at MSRA, March 10, 2021.
One paper is accepted by CVPR'21.
Two papers are accepted by ICRA'21.
Two papers are accepted by ICLR'21.
I gave a talk at Nvidia, Nov, 2020.
One papers is accepted by NeurIPS'20.
One papers is accepted by ECCV'20.
One papers is accepted by ICML'20 (oral).