Yixiang Zhu

Papers

Total Citations

H-Index

About

Yixiang Zhu is a researcher advancing the field of autonomous vehicle localization through innovative 2D-3D registration techniques. Their primary research focuses on fusing image and LiDAR point cloud data to achieve robust, end-to-end robot localization—a critical capability for accurate navigation and mobile manipulation in real-world environments. Zhu’s major contribution lies in developing a unified framework that bridges the gap between economical image-based sensors and illumination-invariant LiDAR mapping, enabling vehicles to localize precisely within pre-built maps without relying on expensive or fragile sensor suites. Their most-cited work, "End-to-End 2D-3D Registration Between Image and LiDAR Point Cloud for Vehicle Localization" (2025), has already garnered 10 citations, demonstrating growing influence in the robotics and autonomous systems community. An earlier version of this work (2023) laid the foundational methodology, accumulating 3 citations. By tackling the core challenge of cross-modal registration, Zhu’s research directly supports practical applications in self-driving cars, warehouse robots, and mobile manipulators. Their work stands out for its end-to-end learning approach, which streamlines the localization pipeline and reduces reliance on handcrafted features, marking a significant step toward more reliable and cost-effective autonomous navigation.