Haoyuan Zhang
Papers
7
Total Citations
120
H-Index
6
About
Haoyuan Zhang is a leading researcher in autonomous robotics, specializing in robust perception, localization, and mapping for mobile robots operating in challenging, real-world environments. His work addresses critical failures in traditional LiDAR- and visual-based SLAM systems under adverse conditions such as rain, snow, smoke, and fog. Zhang’s most impactful contribution is the introduction of **NTU4DRadLM**, the first 4D radar-centric multi-modal dataset for localization and mapping, which has already garnered 33 citations and is enabling a new wave of all-weather SLAM research. He has also made seminal advances in **global localization within geometrically repetitive and ambiguous environments**—such as office corridors and warehouses—where GNSS is unavailable and traditional methods fail. His work on infrastructure-free, magnetic field-aided localization (16 and 11 citations) provides a flexible, cost-effective solution for autonomous vehicles in these settings. Additionally, Zhang has pioneered day-and-night collaborative dynamic mapping using multimodal sensors (42 citations) and developed novel deep-learning stereo matching networks that adapt to low-light and night scenes. His research consistently pushes the boundaries of robot autonomy, ensuring reliable operation across diverse and demanding conditions.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Global Localization in Repetitive and Ambiguous Environments16 citations · 2023
- 4
- 5Soft Warping Based Unsupervised Domain Adaptation for Stereo Matching10 citations · 2021
- 6Magnetic Field-Aided Global Localization in Repetitive Environments6 citations · 2023
- 7