Sirui Chen
Papers
4
Total Citations
32
H-Index
3
About
Sirui Chen is a robotics researcher whose work lies at the intersection of simulation, control, and human-robot interaction, with a particular focus on deformable object manipulation and model-based reinforcement learning. Chen’s most impactful contribution is **DiffSRL**, a framework that learns dynamical state representations for deformable objects using differentiable simulation—enabling robots to understand complex material behaviors and constraints directly from data. This work has garnered 12 citations and represents a foundational step toward making robot manipulation of soft, non-rigid objects both tractable and sample-efficient. Chen further advanced the field with a real-time model predictive control and system identification method (10 citations), which allows robots to continuously improve their models and controllers even after deployment, bridging the simulation-to-reality gap. More recently, Chen introduced **ARCap**, a system that uses augmented reality feedback to collect high-quality human demonstrations for robot learning, and developed a one-shot transfer method for long-horizon extrinsic manipulation tasks. These contributions demonstrate a consistent thread: enabling robots to learn complex, contact-rich behaviors from limited data while leveraging both simulation and human guidance. Chen’s work is shaping how robots perceive and interact with the physical world.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4