Wendong Mao

Sun Yat-sen University, Nanjing University

Papers

Total Citations

H-Index

About

Wendong Mao is a rising researcher at the forefront of efficient hardware design for artificial intelligence, specializing in algorithm-hardware co-optimization for deep neural networks (DNNs) and 3D point cloud processing. His major contributions lie in developing specialized accelerators that bridge the gap between computationally demanding AI models and real-time, energy-constrained applications like autonomous driving, augmented/virtual reality (AR/VR), and intelligent robotics. In his highly cited tutorial on sparse DNN inference and training (2023, 7 citations), Mao provides a comprehensive guide to hardware accelerator design for model pruning, a critical technique for compressing memory-intensive networks. He further advances the field with a low-latency framework for 3D point cloud processing (2023, 5 citations), directly tackling the computational inefficiency caused by data sparsity. Most recently, his work on a precision-scalable vision accelerator for robotic depth estimation (2024, 3 citations) demonstrates his commitment to enabling real-time, autonomous decision-making. Through these contributions, Mao is shaping the future of efficient edge AI, making him a key figure to watch for students and researchers interested in the intersection of hardware architecture and intelligent systems.