Tao Kong

Papers

Total Citations

H-Index

About

Tao Kong is an emerging robotics and computer vision researcher whose work bridges the gap between perception and physical interaction in autonomous systems. His research spans two compelling frontiers: legged robot locomotion and robotic manipulation, with a particular focus on enabling machines to better understand and navigate their environments. In his 2025 work on world model-based perception for visual legged locomotion, Kong addresses a fundamental challenge in robotics — the data inefficiency of learning directly from high-dimensional visual inputs. By leveraging world models, his approach provides robots with a more structured understanding of terrain and surroundings, combining proprioception and vision to enable robust movement across varied surfaces. This work has already attracted 5 citations since its publication. His 2021 contribution on simultaneous semantic and collision learning for 6-DoF grasp pose estimation tackles the notoriously difficult problem of robotic grasping in cluttered environments. Rather than relying on multi-stage pipelines or pre-known object geometry, Kong's method integrates scene understanding and collision awareness in a unified framework, representing a meaningful step toward more practical robotic manipulation systems. Though early in his citation trajectory, Kong's interdisciplinary approach positions him as a promising voice in intelligent robotics research.