Iām a 2nd year Ph.D student at Computer Science & Engineering Department, UC San Diego, where I am honored to be advised by Prof. Hao Su.
I received my Bachelor degree at Department of Physics, Peking University, where I am fortunate to be advised Prof. He Wang.
I'm building dexterous robots that can learn like we do. š¤ I'm recently interested in using human demonstrations, captured via teleoperation and hand tracking,
to teach robots complex manipulation skills. My long-term goal is to make them an essential part in our daily lives.
I'm open to collaborations, feel free to contact me!
Publications are listed by recency. Key papers are in the Selected tab.
Learning Adaptive Dexterous Grasping from Single Demonstrations
Liangzhi Shi*,
Yulin Liu*,
Lingqi Zeng*,
Bo Ai,
Zhengdong Hong,
Hao Su† Website
/
arXiv
/
video
IROS 2025International Conference on Intelligent Robots and Systems
Oral Presentation
TL;DR:
AdaDexGrasp learns grasping skills from single demonstrations and uses a vision-language model to adaptively select the correct one based on user instructions.
RSS 2025Robotics: Science and Systems
Oral Presentation @ ICLR-Robot Learning
TL;DR:
ManiSkill3 is an open-source, GPU-parallelized robotics simulator focusing on generalizable manipulation with contact-rich physics.
SpaRP: Fast 3D Object Reconstruction and Pose Estimation from Sparse Views
Chao Xu,
Ang Li,
Linghao Chen,
Yulin Liu,
Ruoxi Shi,
Hao Su†,
Minghua Liu† Website
/
arXiv
ECCV 2024European Conference on Computer Vision
Oral Presentation @ ECCV-CV4Metaverse
TL;DR:
Given sparse unposed views, we leverage rich priors embedded in multiview diffusion models to predict their poses and reconstruct the 3D shape.
3D-Adapter: Geometry-Consistent Multi-View Diffusion for High-Quality 3D Generation
Hansheng Chen,
Bokui Shen,
Yulin Liu,
Ruoxi Shi,
Linqi Zhou,
Connor Z. Lin,
Jiayuan Gu,
Hao Su,
Gordon Wetzstein,
Leonidas Guibas† Website
/
arXiv
/
code
arXiv 2024
TL;DR:
3D-Adapter enables high-quality 3D generation using a 3D feedback module attached to a base image diffusion model for enhanced geometry consistency.
Delving into Discrete Normalizing Flows on SO(3) Manifold for Probabilistic Rotation Modeling
Yulin Liu*,
Haoran Liu*,
Yingda Yin*,
Yang Wang,
Baoquan Chen†,
He Wang† Website
/
arXiv
/
code
/
video
CVPR 2023Computer Vision and Pattern Recognition Conference
TL;DR:
We propose a discrete normalizing flow on SO(3) manifold, through which one can not only effectively express arbitrary distributions on SO(3),
but also conditionally build the target distribution given input observations.
Selected Honors
Wu-Si Scholarship, highest scholarship in Peking University2022
National Scholarship (Top 1%), Ministry of Education in China2021
Merit Students (Every year), Peking University2019-2023
First-Class Scholarship, Peking University2020
1st Prize in Chinese Physics Olympiad, Ministry of Education in China2018