Papers
7
Total Citations
424
H-Index
5
About
Gangwei Xu is a rising star in computer vision, whose research centers on a critical challenge: enabling machines to perceive depth accurately and efficiently through stereo matching. His major contributions have fundamentally rethought how cost volumes—the core representation for matching pixels between stereo images—are constructed. Xu’s seminal work, the “Attention Concatenation Volume” (ACV), introduced a novel method that generates more informative and concise cost volumes by leveraging attention mechanisms, achieving state-of-the-art accuracy while maintaining high efficiency. This breakthrough paper has garnered over 269 citations, underscoring its profound impact on the field. Building on this foundation, Xu extended his work to tackle persistent challenges like matching ambiguities in ill-posed regions and large disparities, culminating in the “Iterative Multi-Range Geometry Encoding Volumes” (IGEV++). This architecture, published in 2025, represents a new frontier in robust stereo matching. His research, consistently published in top venues, is essential reading for anyone working on autonomous driving, robotics, or 3D scene understanding, demonstrating a clear trajectory from foundational innovation to advanced system design.
Research Focus
Key Achievements
Top Papers
- 1Attention Concatenation Volume for Accurate and Efficient Stereo Matching269 citations · 2022
- 2Accurate and Efficient Stereo Matching via Attention Concatenation Volume84 citations · 2023
- 3IGEV++: Iterative Multi-Range Geometry Encoding Volumes for Stereo Matching37 citations · 2025
- 4
- 5Attention Concatenation Volume for Accurate and Efficient Stereo Matching10 citations · 2022
- 6
- 7