Qiang Lin
Papers
1
Total Citations
8
H-Index
1
About
Qiang Lin is a leading researcher in computer architecture and hardware acceleration for deep learning, with a focus on efficient 3D convolutional neural network (CNN) implementations. His major contribution lies in bridging the gap between computationally intensive 3D CNNs and real-time deployment on edge platforms. Lin's most cited work, "A-U3D: A Unified 2D/3D CNN Accelerator on the Versal Platform for Disparity Estimation" (2022, 8 citations), introduces a novel hardware accelerator that efficiently handles both 2D and 3D convolutions for disparity estimation—a critical task in autonomous driving and robotics. By leveraging the Versal adaptive compute acceleration platform, Lin addresses the significant computational overhead introduced by the disparity dimension in 3D CNNs, achieving practical performance for real-world applications. His work is notable for its unified architecture that maximizes hardware utilization while maintaining accuracy, making it a key reference for researchers developing embedded AI systems. Lin's research continues to push the boundaries of efficient deep learning inference, enabling smarter and more responsive autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1