Junyu Wei

National University of Defense Technology

Papers

Total Citations

H-Index

About

Junyu Wei is a rising researcher in computer vision and robotics, whose work focuses on advancing semantic understanding and spatial intelligence in dynamic environments. His most cited contributions center on two critical challenges: RGBD semantic segmentation and robust SLAM in non-static scenes. In his 2022 paper "Link-RGBD," Wei introduced a cross-guided feature fusion network that significantly improves how depth information is leveraged alongside RGB data for pixel-level scene parsing, addressing a long-standing bottleneck in the field. This work has already garnered 24 citations, reflecting its impact on the segmentation community. Complementing this, his "SOLO-SLAM" paper tackles the fragility of traditional SLAM systems in dynamic real-world settings by proposing a parallel semantic SLAM algorithm that isolates moving objects from static map construction. With 19 citations, this work demonstrates Wei’s ability to bridge perception and localization—a key requirement for autonomous robots. Together, these contributions showcase his talent for designing modular, task-specific neural architectures that push the boundaries of embodied AI, making him a promising voice in the next generation of vision-and-robotics researchers.