Papers
4
Total Citations
35
H-Index
4
About
Chao Zhao is a robotics researcher whose work lies at the intersection of embodied intelligence, dexterous manipulation, and deformable object handling. His key contributions span three critical areas: long-horizon task reasoning, data-driven grasping in clutter, and manipulation of highly deformable objects like paper. Zhao introduced ERRA, an embodied architecture that tightly couples probabilistic inference to enable robots to jointly reason, plan, and interact for complex language-conditioned manipulation tasks (15 citations). In bin picking, he pioneered a "dig-grasping" approach that leverages data-driven physical interaction—digging through clutter before grasping—rather than relying solely on direct pinch grasps, significantly improving success rates in dense clutter (12 citations). His work on FlipBot tackles the notoriously difficult problem of singulating and grasping paper-like objects through coarse-to-fine exteroceptive-proprioceptive exploration (4 citations). Zhao also contributed to deep dexterous grasping from single views, creating a simulator and a dataset of 2.4 million grasps to enable novel object manipulation with multi-fingered hands (4 citations). His research demonstrates a consistent focus on bridging perception, physical interaction, and reasoning for real-world robotic manipulation.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Deep Dexterous Grasping of Novel Objects From a Single View4 citations · 2022