Zhaoxiang Zhang
Papers
3
Total Citations
127
H-Index
3
About
Zhaoxiang Zhang is a versatile researcher whose work spans computer vision, deep learning, and neural computation, with particular emphasis on 3D perception and generative AI. His most recognized contribution, "SARPNET: Shape Attention Regional Proposal Network for LiDAR-based 3D Object Detection" (2019), has garnered 99 citations and represents a significant advancement in autonomous driving perception, introducing shape-aware attention mechanisms to improve the accuracy and robustness of 3D object detection from point cloud data. More recently, Zhang has extended his expertise into the rapidly evolving domain of generative models, with his 2024 work "StableMoFusion" addressing critical challenges in diffusion-based human motion generation — specifically targeting robustness and computational efficiency, areas that remain open problems in the field. His earlier work on primal neural networks for online quadratic programming (2017) reveals a strong mathematical foundation underpinning his applied research. Across these contributions, Zhang demonstrates a rare ability to bridge theoretical rigor with practical innovation, making meaningful impacts across 3D sensing, human motion synthesis, and optimization. His growing citation record reflects increasing recognition within both the computer vision and machine learning communities.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3