Xuesheng Bian
Papers
2
Total Citations
17
H-Index
2
About
Xuesheng Bian is a researcher advancing the frontier of 2D-3D cross-domain feature learning, with a primary focus on computer vision and point cloud processing. His work addresses the fundamental challenge of establishing reliable correspondences between 2D images and 3D point clouds—a critical capability for applications in robotics, augmented reality, and 3D reconstruction. Bian’s most notable contribution is the development of 2D3D-MVPNet, a novel framework that leverages multi-view projections of point clouds to learn robust cross-domain feature descriptors. This work, which has garnered 14 citations, introduces a sophisticated approach to bridging the representational gap between 2D and 3D data. His earlier research on hard triplet loss and spatial transformer networks for cross-domain descriptors further demonstrates his commitment to improving matching accuracy under challenging conditions. Bian’s contributions are particularly impactful for tasks requiring precise alignment between visual and geometric data, and his methods have been recognized for their potential to enhance autonomous navigation and scene understanding systems. With a growing citation record, he is establishing himself as a promising voice in the field of 3D vision and multimodal learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2