CV

Education

• MPhil in Robotics and Autonomous Systems, The Hong Kong University of Science and Technology (Guangzhou), 2025–Present. Supervisor: Prof. Haoang Li.
• B.S. in Robotics Engineering, Harbin Institute of Technology, 2021–2025.

Research and work experience

• Algorithm Intern, 深圳超维动力科技有限公司
  • Contributed to embodied AI foundation model design and training-data development, including harmonizing action spaces and velocity representations across multiple data sources.
  • Expanded background-texture diversity in training data to improve generalization in visually complex environments.
• Algorithm Engineer Intern, Yijiahe Technology Co., Ltd., Nanjing, Jan–Apr 2023
  • Developed vision-guided control and real-time trajectory planning pipelines with OpenCV, Eigen, PCL, MoveIt, and ROS.
  • Worked on hand-eye calibration, humanoid manipulator simulation, and assembly-line manipulation.

Research focus

• Embodied AI, robot foundation models, and vision-language-action models.
• World models, WAM pre-training, robot data processing, and long-horizon manipulation.
• Memory-dependent execution and reliable evaluation on real robots.

Publications

• RoboMemArena: A Keyframe-Annotated Benchmark for Memory-Dependent Robotic Manipulation, preprint, 2026.
• Static-Dynamic Disentanglement for Efficient Multi-Frame Vision-Language-Action Models, NeurIPS 2026, Poster.
• EnvTrap: Revealing the Environment-Only Attack Surface in Embodied AI via Consequence-Blind Action Execution, NeurIPS 2026, Poster.
• HCSG: Human-Centric Semantic-Geometric Reasoning for Vision-Language Navigation, preprint, 2026.
• Rethinking the Practicality of Vision-Language-Action Model: A Comprehensive Benchmark and An Improved Baseline, ICRA 2026.
• High-Accuracy Early Recognition of Upper-Limb Motions for Exoskeleton-Assisted Mirror Rehabilitation, IEEE RA-L, 2025.
• MonoAttack: A Strong Attack Framework with Depth-Migration and Attribute-Tampering for Monocular 3D Object Detection, IJCNN 2025.
• Manipulating the Bounding Box: Multimodal Controlled Backdoor Attacks on 3D Visual Grounding Models, IJCNN 2025.
• Exploring Disentangled Appearance-Motion Contexts for Temporal Activity Localization, IJCNN 2025.

Technical skills

Python, C++, PyTorch, ROS/ROS2, MATLAB, Bash, TensorFlow, JAX, Stable Baselines3, RLlib, Gazebo, Isaac Sim, MuJoCo, PyBullet, CUDA, OpenCV, MoveIt, and point-cloud processing.

Awards

• International Second Prize, U.S. College Student Mathematical Modeling Competition.
• National Second Prize, Offshore Engineering Equipment Competition.
• National Ranking 32/3809, IKCEST Big Data Competition.