Xinlin Yang
Papers
2
Total Citations
29
H-Index
2
About
Xinlin Yang is a leading researcher in the field of efficient deep learning acceleration, with a primary focus on reconfigurable computing and hardware architecture for convolutional neural networks (CNNs). His work is centered on overcoming the critical computation and I/O bottlenecks that limit the performance of modern, high-resolution neural networks. Yang’s major contributions include the development of novel accelerator designs on Xilinx’s advanced Versal platform. His most cited work, "XVDPU: A High Performance CNN Accelerator on the Versal Platform Powered by the AI Engine" (2022, 21 citations), demonstrates a pioneering approach to leveraging the platform’s AI Engine for high-throughput 2D CNN inference. He further extended this architecture with "A-U3D: A Unified 2D/3D CNN Accelerator on the Versal Platform for Disparity Estimation" (2022, 8 citations), a significant achievement that tackles the computationally dense demands of 3D CNNs for critical applications like autonomous driving and robotics. By creating a unified solution for both 2D and 3D processing, Yang’s research directly addresses the need for practical, high-performance hardware in real-world computer vision systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2