Papers
8
Total Citations
39
H-Index
5
About
Haodong Zhang is a robotics researcher whose work spans robot learning, autonomous navigation, and humanoid locomotion — fields where intelligence and physical capability converge. His most recognized contribution, "Learning to Fill the Seam by Vision" (2022, 12 citations), demonstrates an innovative approach to sub-millimeter peg-in-hole assembly by mimicking human visual attention, enabling robots to handle unseen shapes in real-world settings — a significant advance for precision manufacturing automation. His broader portfolio reveals a researcher comfortable bridging perception, planning, and control: he has developed deep reinforcement learning strategies for brachiation robots (5 citations), model-based navigation frameworks for crowded pedestrian environments (5 citations), and adaptive computer vision pipelines for industrial inspection robots (6 citations). More recently, Zhang has turned his attention to humanoid robotics, proposing a Generative Motion Prior framework for naturalistic locomotion and a whole-body motion imitation system that reconciles the kinematic gap between humans and full-size humanoid platforms. With work spanning path planning, laser-based metrology, and learned motion priors, Zhang represents a versatile voice in modern robotics research whose contributions are increasingly shaping how robots move, perceive, and interact in complex, human-centered environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5Learning World Transition Model for Socially Aware Robot Navigation5 citations · 2020
- 6Natural Humanoid Robot Locomotion with Generative Motion Prior2 citations · 2025
- 7
- 8