Huawei Li

Chinese Academy of Sciences, Institute of Computing Technology, Shanghai University

Papers

Total Citations

258

H-Index

About

Huawei Li is a researcher whose work sits at the compelling intersection of deep learning, hardware acceleration, and embedded systems design. He is best known for his pioneering contributions to neural network accelerator design, most notably through the influential **DeepBurning** framework (2016), which has garnered over 208 citations and addressed the critical challenge of building energy-efficient hardware solutions for convolutional and artificial neural networks in embedded vision, robotics, and cyberphysical systems. Complementing this, his work on re-architecting on-chip memory subsystems for machine learning accelerators demonstrated innovative approaches to managing the demanding storage requirements of deep CNNs on resource-constrained devices. More recently, Li has expanded his research focus to encompass the reliability and safety of deployed deep learning systems, publishing on fault-tolerant deep learning from a hierarchical perspective — a timely contribution as AI increasingly powers safety-critical applications like autonomous driving. His earlier work in robotic kinematics analysis further reflects his broad engineering foundations. Across his career, Li has consistently tackled the practical barriers between cutting-edge deep learning algorithms and their real-world deployment, making him a meaningful contributor to embedded AI and dependable computing research.