Robot Learning Embodied AI Dexterous Manipulation

Jianhan (Jaime) Ma 马健涵

I am currently a research engineer at Noematrix. Previously I was a researcher at Shanghai Qi Zhi Institute, under the mentorship of Prof. Huazhe Xu at Tsinghua University. I obtained my Master of Science degree at the Electrical & Computer Engineering department of UC San Diego, where I was fortunate to be advised by Prof. Xiaolong Wang.

My research is focused on advanced robot learning techniques to acquire generalized skills for locomotion & manipulation. My passion lies in exploring how distinct yet interconnected modalities, including visual and language inputs, can synergistically enhance the capabilities of intelligent systems.

Before joining UCSD, I obtained my Bachelor of Engineering (B.E.) degree in Electrical Engineering and its Automation at Zhejiang University, supervised by Prof. Liangjing Yang and Prof. Zuozhu Liu.

Publications & Preprints

* Core Contributors † Project Lead

UniDex: A Robot Foundation Suite for Universal Dexterous Hand Control from Egocentric Human Videos

Gu Zhang*^†, Qicheng Xu*, Haozhe Zhang*, Jianhan Ma*, Long He*, Yiming Bao*, Zeyu Ping, Zhecheng Yuan, Chenhao Lu, Chengbo Yuan, Tianhai Liang, Xiaoyu Tian, Maanping Shao, Feihong Zhang, Mingyu Ding, Yang Gao, Hao Zhao, Hang Zhao, Huazhe Xu

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026

Project Page arXiv Code Model

UniDex is a robot foundation suite to construct large-scale robot data from egocentric human videos, and a vision-language-action policy for universal dexterous hand control.

Generalized Animal Imitator project preview

Generalized Animal Imitator: Agile Locomotion with Versatile Motion Prior

Ruihan Yang*, Zhuoqun Chen*, Jianhan Ma*, Chongyi Zheng*, Yiyu Chen, Quan Nguyen, Xiaolong Wang

Conference on Robot Learning (CoRL), 2024

CoRL Deployable Workshop, 2023 Workshop Best Paper Award

Project Page arXiv Video

Our system learns a Single Instructable Motion Prior from a diverse reference motion dataset.

Selected Projects

Camera-based Augmented Reality Endoscope Auxiliary System

Advisor: Prof. Liangjing Yang

Project Report

We engineered a sophisticated Augmented Reality (AR) auxiliary system designed to accurately track and tag tumor locations in real-time. This system was adeptly integrated with the vision system of an endoscope mounted on an OpenManipulator-X robot arm. It effectively displayed the pinpointed tumor locations directly onto the endoscope's visual feed, enhancing precision and guidance during medical procedures.

Self-supervised Learning for Semantic Tooth Segmentation on Cone-Beam CT Image

Advisor: Prof. Zuozhu Liu

Project Report

We proposed an innovative self-supervised representation learning strategy to enhance the accuracy of semantic segmentation in Cone-Beam Computed Tomography (CBCT) datasets.