Zhiwei Fang

Hunan Institute of Science and Technology

Papers

Total Citations

H-Index

About

Zhiwei Fang is a researcher advancing the frontier of 3D geometric deep learning, with a primary focus on point cloud analysis for autonomous driving, robotic perception, and remote sensing. His most notable contribution is the development of PointGA, a lightweight Transformer-based model introduced in his 2025 paper "Geometrically aware transformer for point cloud analysis" (9 citations). This work addresses a critical challenge in the field: enabling efficient and accurate geometric perception from sparse 3D point cloud data. By enhancing the Transformer architecture with geometric awareness, Fang's model achieves superior performance while maintaining computational efficiency—a key requirement for real-time applications in autonomous systems. His research sits at the intersection of computer vision, geometric deep learning, and efficient neural architecture design, demonstrating how attention mechanisms can be adapted to capture spatial relationships in unstructured 3D data. Fang's work is particularly impactful for students and researchers working on 3D perception tasks, offering a practical solution that balances accuracy with deployability. As the demand for robust 3D understanding grows across robotics and autonomous navigation, his contributions provide a foundation for future innovations in geometrically-aware learning systems.