Xiaoliang Sun

National University of Defense Technology

Papers

Total Citations

H-Index

About

Xiaoliang Sun is a computer vision researcher whose work focuses on advancing monocular 3D object pose tracking for real-world robotic applications. His primary research areas include robust pose estimation, visual tracking under challenging conditions, and vision-based robotic manipulation. Sun’s major contributions address two critical limitations in existing pose-tracking methods: handling large interframe pose shifts and adapting to extreme scale variations. His 2022 paper on robust monocular pose tracking for large pose shifts (18 citations) pioneered techniques that break free from the traditional motion continuity assumption, enabling accurate tracking even during rapid, discontinuous movements. In parallel, his work on scale-adaptive region-based methods (3 citations) specifically targets robotic manipulation scenarios where camera-object distance varies dramatically, causing significant scale changes in image sequences. This research directly impacts practical robotics by allowing manipulators to maintain precise 6-degree-of-freedom tracking without requiring expensive depth sensors. While still early in his career, Sun’s focused contributions to solving fundamental robustness problems in monocular tracking—particularly for industrial and service robotics—demonstrate promising potential. His work bridges the gap between theoretical pose estimation and the demanding requirements of real-world robotic systems operating in unstructured environments.