Papers
3
Total Citations
22
H-Index
3
About
Wenhui Wei is a researcher advancing the frontier of visual odometry and SLAM for robotics, with a focus on making autonomous systems more robust in dynamic, real-world environments. Wei’s key contributions lie in developing lightweight, self-supervised methods that enhance geometric and semantic feature extraction—critical for robots navigating unpredictable spaces. Their most-cited work, "GSL-VO" (2023, 10 citations), introduces a geometric-semantic enhanced framework that significantly improves perception in unseen dynamic environments, addressing a core limitation of learning-based visual odometry. Building on this, "Fine-MVO" (2024, 8 citations) tackles coarse feature representations by enabling fine-grained feature enhancement without reliance on costly labels, a breakthrough for autonomous driving and indoor robotics. Most recently, "BotVIO" (2025, 4 citations) integrates transformer architectures into visual-inertial odometry, offering a lightweight yet powerful solution for robust localization. Despite the early stage of these publications, Wei’s work is already shaping the next generation of efficient, label-free navigation systems. Their research is particularly notable for balancing computational efficiency with high accuracy—a crucial trade-off for deploying SLAM on resource-constrained robotic platforms.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3