Xinge Zhao

Wuhan University, Northeastern University

Papers

Total Citations

H-Index

About

Xinge Zhao is an emerging researcher specializing in computer vision, autonomous navigation, and robotic perception, with a particular focus on Visual Place Recognition (VPR) and Simultaneous Localization and Mapping (SLAM) systems. Their work addresses critical challenges in enabling robots and autonomous vehicles to navigate reliably across diverse and dynamically changing environments. Zhao's most notable contribution extends the widely-used ORB-SLAM3 framework to support 360-degree panoramic video, tackling a significant gap in mainstream monocular SLAM systems that previously offered insufficient support for omnidirectional data. This 2022 work demonstrates a practical, highly generalizable approach to autonomous positioning and orientation. More recently, Zhao has advanced the field of VPR by developing methods that overcome the persistent challenges of appearance changes, varying illumination, and perceptual aliasing. Their 2024 papers introduce multi-modal feature fusion with time-constrained graph attention aggregation and neighborhood consensus guided matching with spatial-channel embeddings — both representing sophisticated solutions for robust scene recognition under real-world conditions. Although early in their research career with a combined citation count across key publications, Zhao's targeted contributions to autonomous driving and mobile robotics position them as a promising voice in an increasingly vital area of artificial intelligence research.