Lixin Xu

Robotics Researcher | Ph.D. Student in Robotics

Guangzhou, CN.

About

I am deeply committed to advancing the field of robotics with a focus on the integration of cognitive and athletic intelligence to create robotics foundation models that could revolutionize our modern industry and society. I am currently a PhD student at the Hong Kong University of Science and Technology (Guangzhou), supervised by Prof. Renjing Xu. Previously, I was a visiting intern in Prof. Yilun Chen's group at Tsinghua AIR, and in Prof. Lin Shao's group in National University of Singapore, focusing on dexterous policy learning and sim-to-real deployment. I earned my Master's degree in ECE at Georgia Tech, and Bachelor's in Mechanical Engineering at Qingdao University.

Work

HKUST (Guangzhou)

|

Research Assistant

→

Summary

Research Assistant in Prof. Renjing Xu's group.

Tsinghua University

|

Research Intern

Beijing, , China

→

Summary

Research Intern in Prof. Yilun Chen's group at Tsinghua AIR, focusing on dexterous policy learning and sim-to-real deployment.

Highlights

Engaged in pioneering research within the Institute for AI Industry Research (AIR), collaborating with Supervisor Yilun Chen and Wenchao Ding.

Applied advanced AI methodologies to complex robotics challenges, contributing to novel solutions and research outcomes.

National University of Singapore, School of Computing

|

Research Intern

Singapore, , Singapore

→

Summary

Contributed to research projects focused on computational aspects of robotics and artificial intelligence under academic guidance.

Highlights

Collaborated with Prof. Lin Shao at the School of Computing, focusing on innovative research in robotics and AI.

Developed and implemented computational models to address specific research objectives, enhancing project progress.

Education

HKUST (Guangzhou)

Guangzhou, , China
→
Ph.D Student Robotics Grade:

Georgia Institute of Technology

Atlanta, Georgia, United States of America
→
M.Sc. Electrical and Computer Engineering Grade: 4.0/4.0

Courses

ECE 6258 Digital Image Processing

ECE 6122 Advanced Programming Techniques

ECE 6550 Linear Systems and Control

CS 7641 Machine Learning

CS 7643 Deep Learning

CS 7638 AI for Robotics

Qingdao University

Qingdao, , China
→
B.Sc. Mechanical Engineering Grade: 90.96/100 (Top 1%)

Awards

National Scholarship

Ministry of Education of the People's Republic of China

Awarded for outstanding academic achievement as part of the top 1% of the graduating class.

National Scholarship

Ministry of Education of the People's Republic of China

Nation Scholarship

Special Prize

6th National Engineering Training Competition

Recognized for exceptional performance in the Material Handling Robot competition, demonstrating superior engineering and problem-solving skills.

Publications

WHIRL: Intervention-Aware World Models with Real-World RL for Dexterous Manipulation

CoRL 2026

Jiaju Yin, Zhenhui Zhang, Lixin Xu, Heng Zhang, Jun Shao, Yating Feng, Arash Ajoudani, Renjing Xu

What Bimanuals Can Do (WBCD) Challenge 2026 - Champion

ICRA 2026 Challenge

Boyang Cai, Jinru Song, Xuanyu Lai, Xiangzheng Wang, Xiangyu Chen, Jiawei Xia, Jiawei Li, Wenxuan Song, Yu Yan, Wenjie Zhang, Qiwei Wu, Yan Qin, Lixin Xu, Weisheng Xu, Yihang Kang, Qiwei Liang, Renjing Xu

PDF-HR: Pose Distance Fields for Humanoid Robots

arXiv

Yi Gu, Yukang Gao, Yangchen Zhou, Xingyu Chen, Yixiao Feng, Mingle Zhao, Yunyang Mo, Zhaorui Wang, Lixin Xu, Renjing Xu

DexFormer: Cross-Embodied Dexterous Manipulation via History-Conditioned Transformer

arXiv

Ke Zhang, Lixin Xu, Chengyi Song, Junzhe Xu, Xiaoyi Lin, Zeyu Jiang, Renjing Xu

DexFlow: A Unified Approach for Dexterous Hand Pose Retargeting and Interaction

arXiv

Xiaoyi Lin, Kunpeng Yao, Lixin Xu, Xueqiang Wang, Xuetao Li, Yuchen Wang, Miao Li

What Bimanuals Can Do (WBCD) Challenge 2025 - Champion

ICRA 2025 Challenge Report

Weize Li, Zhengxiao Han, Lixin Xu, Xiangyu Chen, Harrison Bounds, Chenrui Zhang, Yifan Xu

DexSinGrap: Learning a Unified Policy for Dexterous Object Singulation and Grasping in Densely Cluttered Environments

R-AL

Lixin Xu, Zixuan Liu, Zhewei Gui, Jingxiang Guo, Zeyu Jiang, Zhixuan Xu, Chongkai Gao, Lin Shao

Languages

Mandarin

English

Cantonese

German

Japanese

Skills

Programming & Tools

Python, PyTorch, C/C++, Isaac Gym, MATLAB/Simulink, ROS, Cmake, LaTeX, Solidworks.

ML & Decision Making

Imitation Learning, Reinforcement Learning, LLM/VLM, Deep Learning, Machine Learning.

Control & Actuation

Nonlinear control, Disturbance rejection control, Dynamics modeling, Linear Systems and Control.

Computer Vision

Digital Image Processing, Monocular Depth Estimation, Video Inpainting, Stereo Vision Odometry, Object Recognition.

Robotics Frameworks

ROS (Robot Operating System).

Projects

ICRA WBCD Challenge

→

Summary

Participated in the ICRA WBCD Challenge, a prestigious robotics competition, achieving recognition for innovative solutions in bimanual manipulation.

Monocular Depth Estimation - FADE Ain't Depth Estimation

→

Summary

A deep learning project focused on enhancing monocular depth estimation using large-scale networks and advanced boosting methods.

C++ Simulation, Games and Visualization

→

Summary

Engaged in multiple C++ projects focusing on simulation, game development, and visualization, showcasing expertise in OpenGL and SFML.

An Attention-Based Video Inpainting Technique for Wire-Removal Scenarios

→

Summary

Developed an innovative autoencoder-based video inpainting model specifically designed for wire-removal in cinematic applications.

An ORB-based Stereo Vision Odometry (Undergrad thesis)

→

Summary

Designed and implemented PyVO, a Python-based visual stereo odometer, as part of an undergraduate thesis project.

The 6th National Engineering Training Competition - Material Handling Robot

→

Summary

Built an STM32-based Mecanum Wheel robot for a national engineering competition, focusing on material handling and object recognition.