Xingwu Ji
Papers
2
Total Citations
62
H-Index
2
About
Xingwu Ji is a leading researcher in autonomous navigation and robotics perception, whose work addresses the critical challenge of reliable positioning in complex, real-world environments. His most impactful contribution is a pioneering adaptive fusion system for GNSS and visual-inertial odometry (VIO), published in 2020 and cited 60 times. This work tackles the persistent problem of intermittent GNSS degradation—a major hurdle for autonomous vehicles—by intelligently fusing sensor data to maintain consistent, accurate global positioning even when satellite signals are lost. More recently, Ji has advanced the field of visual place recognition (VPR) with his 2024 paper on Bird's Eye View Graph Matching (BEVGM). This novel method uses graph-based matching from a bird's-eye perspective to dramatically improve VPR robustness against severe appearance changes, reverse viewpoints, and heterogeneous data—scenarios that traditionally cripple existing systems. By bridging the gap between theoretical sensor fusion and practical, real-world deployment, Ji's research is directly enabling safer, more reliable autonomous navigation. His work stands as a vital resource for engineers and researchers developing next-generation robotics and self-driving technologies.
Research Focus
Key Achievements
Top Papers
- 1
- 2