I’m a robotics researcher and developer focused on intelligent manipulation, skill learning, and human-robot interaction.
I build systems that combine real-time control with learning-based decision-making to make robots more adaptive and capable.
🔗 Check out my personal website:
🌐 Cheng Website
- Designing and learning reusable robotic skills from demonstration, teleoperation, and reinforcement learning
- Building scalable skill libraries and high-level planners for complex tasks
- Developing Cartesian and joint-space planning frameworks
- Integrating with real-time controllers
- Using vision-language models (VLMs) for grounding, task recognition, and semantic skill selection
- Coordinated control and shared autonomy across multi-arm or mobile robot platforms
- Languages: Python, C++
- Frameworks: ROS 2, MuJoCo, V-REP, Rocos-app, Ruckig, PyTorch, OpenCV
- VLMs & Learning: ACT, Diffusion Policy, 3D Diffusion Policy,Transformers
- Platforms: Franka Panda, Jetson, Realsense, Ubuntu Linux
- GitHub: cheng9911
- Website: 🌐Cheng Website