Xiaoheng Li
Papers
1
Total Citations
2
H-Index
1
About
Xiaoheng Li is a rising researcher in computer vision and 3D scene understanding, with a focus on self-supervised learning for point cloud analysis. His most notable contribution is the development of PVFT-Net, a novel point-voxel fusion framework that integrates transformer architectures for self-supervised scene flow estimation. This work, published in 2025, addresses the critical challenge of estimating 3D motion from unstructured point cloud data without requiring expensive ground-truth annotations. By combining the efficiency of voxel-based representations with the precision of point-based methods, Li’s approach achieves robust performance in dynamic environments, such as autonomous driving and robotics. Though early in his career, his work has already garnered attention, with his most-cited paper accumulating 2 citations in a short span. Li’s research pushes the boundaries of self-supervised learning in 3D vision, offering scalable solutions for real-world applications where labeled data is scarce. His innovative fusion of point-voxel methods with transformer mechanisms marks a promising direction for future advancements in scene flow estimation and spatial-temporal reasoning.
Research Focus
Key Achievements
Top Papers
- 1