Xuanheng Li
Papers
2
Total Citations
14
H-Index
2
About
Xuanheng Li is at the forefront of robotic dexterous manipulation, specializing in functional grasp generation, visuo-tactile sensing, and deep reinforcement learning for high-degree-of-freedom (DoF) robotic hands. His major contribution lies in addressing the fundamental challenge of enabling multi-fingered hands to perform complex, contact-rich grasps—a task far more demanding than traditional low-DOF gripper control. Li pioneered the creation of the "DexFuncGrasp" dataset (2024, 12 citations), a cost-effective real-simulation annotation system that provides a critical foundation for training grasp generation models, filling a gap where most existing datasets focus only on stable grasps rather than functional ones. Building on this, his work on "Multi-fingered Hand Grasps with Visuo-Tactile Fusion" (2025, 2 citations) introduces a novel multi-agent deep reinforcement learning framework that synergizes visual and tactile feedback, effectively taming the high-dimensional action space of dexterous hands. This breakthrough promises to advance robotic manipulation in unstructured environments, from manufacturing to assistive robotics. With his innovative datasets and learning algorithms, Li is shaping the next generation of intelligent, dexterous robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2