Yaobo Liang

Microsoft Research Asia (China)

Papers

2

Total Citations

7

H-Index

2

About

Yaobo Liang is at the forefront of embodied AI, pioneering the integration of large vision-language models with robotic control. His research centers on developing foundational models for robotic manipulation, with a particular focus on dexterous grasping and language-guided task execution. Liang’s most notable contribution is **CogACT**, a foundational Vision-Language-Action (VLA) model that synergizes cognition and action, enabling robots to generalize to unseen scenarios through natural language commands. This work has already garnered significant attention, with 5 citations since its 2024 release. Additionally, his **UniGraspTransformer** simplifies the complex, multi-step training pipelines of prior dexterous grasping methods, offering a scalable and efficient alternative for universal robotic grasping. By distilling policy learning into a streamlined Transformer architecture, Liang has made dexterous manipulation more accessible and practical. His work bridges the gap between high-level language understanding and low-level motor control, directly impacting the future of service robotics and industrial automation. With a growing citation footprint and a clear trajectory toward generalist robotic agents, Yaobo Liang is a rising star in the intersection of computer vision, natural language processing, and robotics.

Research Focus

Key Achievements

2
H-Index
2
Papers
7
Total Citations
4
Avg Citations/Paper
🏆 Most Cited Paper
CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
5 citations · 2024
📈 Most Prolific Year: 2024 (1 Papers)
🤝 Key Collaborators: 25
🏛 Institutions: Microsoft Research Asia (China)

Top Papers

  1. 1
  2. 2

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 12 days ago