Xiaocan Li

Shandong Institute of Automation, Chinese Academy of Sciences

Papers

Total Citations

H-Index

About

Xiaocan Li is a robotics researcher whose work sits at the intersection of tactile perception, robot learning, and autonomous manipulation. Li’s most influential contribution is the **Generalized Visual-Tactile Transformer (GVT-Transformer)**, a pioneering deep learning architecture that fuses visual and tactile data to enable robust slip detection during robotic grasping. This work, which has garnered **15 citations**, addresses a long-standing challenge in dexterous manipulation by allowing robots to sense and react to object slippage in real time—a critical capability for tasks ranging from assembly to assistive robotics. Li has also advanced the field of **robot skill acquisition** through an automatic learning system that extracts reusable skills from real-world demonstrations using RGB-D cameras, published in 2019 with 5 citations. This system reduces the programming burden in robotics by enabling robots to learn directly from human or robot demonstrations. Li’s research is notable for its practical, real-world focus, bridging the gap between high-level learning algorithms and low-level sensorimotor control. By combining transformer-based architectures with multimodal sensing, Li is helping to build more adaptive, intelligent robots capable of operating safely in unstructured environments.