Papers
4
Total Citations
61
H-Index
4
About
Hengwang Zhao is a leading researcher in robotics and autonomous systems, specializing in point cloud registration and cross-modal visual localization. His work addresses critical challenges in enabling mobile robots and intelligent vehicles to navigate with high accuracy and low cost. Zhao’s most impactful contribution is the development of the CentroidReg framework, a global-to-local approach for partial point cloud registration that overcomes the sensitivity of previous algorithms to noise and occlusion, earning 20 citations. He further advanced the field with G3DOA, a generalizable 3D descriptor leveraging overlap attention for robust registration in autonomous driving, cited 13 times. In cross-modal localization, Zhao’s 2023 paper on monocular visual localization in prior LiDAR maps using semantic consistency (23 citations) demonstrates a novel method for stable image-to-map alignment. His most recent work, GOGICP, introduces a real-time Gaussian octree-based GICP method for faster point cloud registration, achieving both accuracy and real-time performance. With over 60 total citations and a consistent focus on practical, deployable solutions, Zhao’s research is shaping the future of robot localization and perception in complex environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4