Papers
12
Total Citations
212
H-Index
6
About
Kechun Xu is a leading researcher in robotic manipulation, focusing on the intersection of vision, language, and action for complex, real-world tasks. Their work centers on enabling robots to intelligently interact with cluttered environments, with key contributions in goal-oriented grasping, language-conditioned manipulation, and object rearrangement. Xu’s most cited work, “Efficient Learning of Goal-Oriented Push-Grasping Synergy in Clutter” (84 citations), pioneered a method for robots to learn pre-grasp actions like pushing to enable stable grasps on target objects. This was extended in “A Joint Modeling of Vision-Language-Action for Target-oriented Grasping in Clutter” (50 citations), which unified visual grounding and grasp generation for language-guided tasks. Xu also contributed to dynamic manipulation with “Neural Motion Prediction for In-flight Uneven Object Catching” and to safe autonomy with “Failure-aware Policy Learning for Self-assessable Robotics Tasks.” Their comprehensive 2025 review on humanoid robots has already garnered 26 citations, reflecting their broad impact. Through these works, Xu has advanced the field toward more capable, adaptive, and instruction-following robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Efficient Learning of Goal-Oriented Push-Grasping Synergy in Clutter84 citations · 2021
- 2
- 3A Comprehensive Review of Humanoid Robots26 citations · 2025
- 4
- 5Neural Motion Prediction for In-flight Uneven Object Catching12 citations · 2021
- 6
- 7
- 8
- 9Efficient learning of goal-oriented push-grasping synergy in clutter4 citations · 2021
- 10Failure-aware Policy Learning for Self-assessable Robotics Tasks3 citations · 2023