Hi there 👋
I am Lijian, a Research Assistant at the HCLLab of The Hong Kong University of Science and Technology (Guangzhou). My research focuses on embodied manipulation learning, robot-centric generative interaction, real-world online reinforcement learning, and tactile manipulation.
Previously, I worked as a Research Assistant at ShanghaiTech University MARSLab on SLAM, manipulation data synthesis, teleoperation systems, and reinforcement learning.
Prior to embarking on my research journey, I worked at the Industrial and Commercial Bank of China Software Development Center (Beijing). I obtained my Bachelor’s Degree in Computer Science and Technology from Tianjin University of Finance and Economic.
- Embodied manipulation learning
- Zero-shot humanoid interaction
- Real-world online reinforcement learning
- Tactile manipulation
- Generative video for robot action acquisition
Python PyTorch ROS/ROS2 MuJoCo Isaac Lab SAC PPO SERL Diffusion Flow Matching VLM ViTPose CoTracker LaTeX
Research Assistant | 2025.06 - Present
- Research on embodied manipulation learning, robot-centric generative interaction, and online RL on real robots.
Research Assistant | 2024.04 - 2025.05
- Worked on SLAM, manipulation data synthesis, teleoperation collection, and reinforcement learning.
- Contributed to dataset compression, grasp generation, teleoperation systems, and embodied manipulation projects.
- High-Quality, ROS Compatible Video Encoding and Decoding for High-Definition Datasets. Li Jian, Bowen Xu, and Sören Schwertfeger.
ROBIO 2024 - Gentle Manipulation Policy Learning via Demonstrations from VLM Planned Atomic Skills. Jiayu Zhou, Qiwei Wu, Jian Li, Zhe Chen, Xiaogang Xiong, Renjing Xu.
AAAI 2026 Oral
- Email:
lijianonwork@126.com - GitHub: lijianfromearth