Papers
1
Total Citations
8
H-Index
1
About
Zhexue Ge is a researcher advancing the field of robotics and computer vision, with a primary focus on visual relocalization and semantic scene understanding. His most cited work, "Object-Plane Co-Represented and Graph Propagation-Based Semantic Descriptor for Relocalization" (2022, 8 citations), addresses a fundamental challenge in robotics: enabling autonomous systems to accurately determine their position despite drastic changes in lighting, weather, and viewpoint. Ge’s key contribution lies in developing a novel hybrid approach that combines object-level and plane-level semantic landmarks, overcoming the limitations of appearance-sensitive image features and ambiguous high-level semantic methods. By introducing a graph propagation-based semantic descriptor, he creates a more robust and discriminative representation for topological map matching. This work directly tackles the fragility of traditional relocalization techniques, offering a pathway toward more reliable long-term robot navigation in dynamic environments. Ge’s research is particularly valuable for applications in autonomous driving, service robotics, and augmented reality, where consistent localization under varying conditions is critical. His innovative co-representation strategy marks a significant step toward bridging the gap between low-level features and high-level semantics in spatial AI.
Research Focus
Key Achievements
Top Papers
- 1