Hang Shi
Papers
1
Total Citations
4
H-Index
1
About
Hang Shi is a researcher whose work sits at the intersection of computer vision, geometric deep learning, and spherical data processing. Their most notable contribution is the development of UVS-CNNs, a novel framework for constructing general convolutional neural networks directly on quasi-uniform spherical images. This work, published in 2024 and already garnering 4 citations, addresses a critical challenge in applying CNNs to non-Euclidean domains, enabling more accurate and efficient analysis of omnidirectional and panoramic imagery. By designing convolution operations that respect the spherical geometry without the distortions common to equirectangular projections, Shi's approach has implications for autonomous navigation, 360-degree video understanding, and remote sensing. The research demonstrates a sophisticated grasp of both theoretical geometry and practical network architecture, offering a scalable solution for real-world spherical data. As this field rapidly expands with the proliferation of VR and 360-degree cameras, Shi's foundational work is poised to become a key reference for future studies in spherical deep learning and geometric representation learning.
Research Focus
Key Achievements
Top Papers
- 1