摘要
由于存储空间和功耗的限制,神经网络模型在嵌入式设备上的存储和计算仍然是一个巨大的挑战。模型压缩作为一种有效的解决方法,受到了越来越多研究者的关注。针对卷积神经网络模型进行了研究,分析了模型中存在的冗余信息,并对国内外学者在神经网络模型压缩方面的研究成果整理,从参数剪枝,权重共享和权重矩阵分解等方面总结了神经网络压缩的主要方法。最后针对神经网络模型发展现状及目前面临的若干主要问题进行了讨论,指出了下一步的研究方向。
However,limited by the memory space and power,it is a challenging task to deploy the deep neural network model in embedded system.As an effective solution,model compression has attracted tremendously attention for researchers.At first,this paper introduced the basic deep neural network model and analyzed the redundant computation in the model.Then,it presented the existing methods for model compression from the aspects of parameter pruning,parameter sharing and weight matrix decomposition.Finally,this paper discussed the potential challenges in deep neural network and direction of future research.
作者
曹文龙
芮建武
李敏
Cao Wenlong;Rui Jianwu;Li Min(University of Chinese Academy of Sciences,Beijing 100190,China;Institute of Software,Chinese Academy of Sciences,Beijing 100190,China;General Chips&Basic Software Research Center,Chinese Academy of Sciences,Beijing 100190,China)
出处
《计算机应用研究》
CSCD
北大核心
2019年第3期649-656,共8页
Application Research of Computers
关键词
神经网络
模型压缩
矩阵分解
参数共享
neural network
model compression
matrix decompression
parameter sharing