ICCT 2022 Invited Speaker

Qiang Li, Southwest University of Science and Technology, China
李强, 西南科技大学

http://www.info.swust.edu.cn/index.php?a=index&f=tc&aid=268&id=21

Li Qiang，Ph.D, he received the Ph.D. degree in biomedical engineering from the University of Science and Technology of China in 2008. And he is currently a professor with the School of Information Engineering, Southwest University of Science and Technology. He is a member of the Teaching Steering Committee of Biomedical Engineering in Sichuan Province, a senior member of China Electronics Society / a member of Internet of Things Youth Technical Group of China Electronics Society, etc. With the target of the key & core technology of industrial Internet of Things and intelligent equipment，his research focuses on the intelligent terminal system design and edge computing, Internet of Things reliable transmission with measurement and control, intelligent analysis and processing.

Speech Title: An Energy-Constrained Optimization-Based Structured Pruning Method for Deep Neural Network Compression

Deep neural networks are widely used in modern intelligent applications due to their superior ability to express reality, and these intelligent applications run on highly energyconstrained edge devices. Neural network structured pruning is an efficient method to reduce the energy consumption of neural networks. This paper proposes an energy-constrained optimization-based structured pruning method to compress deep neural network. This method uses the energy budget provided by edge devices as a constraint for neural network pruning. The key idea is to formulate deep neural network structured pruning as an optimization problem, in which the energy estimate of each layer is used as the Frobenius norm optimization constraint for the convolution kernel weights. Then the importance of all convolution kernels in the network can be evaluated, thereby the neural network structure can be pruned according to the importance to reduce the energy consumption and memory size of the neural network. Compared with traditional heuristic structured pruning methods, our proposal enables neural networks to achieve higher accuracy with the same or lower energy budget.