Guangyu Li
Papers
2
Total Citations
16
H-Index
2
About
Guangyu Li is a leading researcher in embodied AI and robot manipulation, whose work centers on advancing multi-embodiment intelligence and scalable learning for robotic systems. His most impactful contribution is the introduction of **RoboMIND** (Multi-embodiment Intelligence Normative Data for Robot Manipulation), a landmark dataset comprising **107,000 demonstration trajectories** across **479 diverse tasks** involving **96 object classes**. Collected via human teleoperation, RoboMIND provides a standardized benchmark for training and evaluating robot manipulation policies across different hardware platforms, addressing a critical gap in generalizable robotic learning. This work has rapidly accumulated **16 citations** since its 2024–2025 release, reflecting its immediate influence on the field. By enabling cross-embodiment skill transfer and reproducible evaluation, Li’s research directly tackles the data scarcity and task diversity challenges that have long hindered real-world robot deployment. His contributions are foundational for researchers working toward foundation models for robotics, offering both a rich dataset and a normative benchmark that accelerates progress toward truly intelligent, adaptable robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2