Convolutional neural network acceleration algorithm based on filters pruning

被引：0
|
作者：
Li H. [1 ]
Zhao W.-J. [1 ]
Han B. [1 ]
机构：
[1] College of Aeronautics and Astronautics, Zhejiang University, Hangzhou
来源：
Zhejiang Daxue Xuebao (Gongxue Ban)/Journal of Zhejiang University (Engineering Science) | 2019年 / 53卷 / 10期
关键词：
Convolutional neural network (CNN); Deep learning; Feature map; Filter; Model compress;
D O I：
10.3785/j.issn.1008-973X.2019.10.017
中图分类号：
学科分类号：
摘要：
A new model acceleration algorithm of convolutional neural network (CNN) was proposed based on filters pruning in order to promote the compression and acceleration of the CNN model. The computational cost could be effectively reduced by calculating the standard deviation of filters in the convolutional layer to measure its importance and pruning filters with less influence on the accuracy of the neural network and its corresponding feature map. The algorithm did not cause the network to be sparsely connected unlike the method of pruning weight value, so there was no need of the support of special sparse convolution libraries. The experimental results based on the CIFAR-10 dataset show that the filters pruning algorithm can accelerate the VGG-16 and ResNet-110 models by more than 30%. Results can be close to or reach the accuracy of the original model by fine-tuning the inherited pre-training parameters. © 2019, Zhejiang University Press. All right reserved.
引用
收藏
页码：1994 / 2002
页数：8
相关论文
共 23 条
  • [1] Krizhevsky A., Sutskever I., Hinton G.E., ImageNet classification with deep convolutional neural networks, International Conference on Neural Information Processing Systems, (2012)
  • [2] Graves A., Schmidhuber J., Framewise phoneme classification with bidirectional LSTM and other neural network architectures, Neural Netw, 18, 5, pp. 602-610, (2005)
  • [3] Szegedy C., Vanhoucke V., Ioffe S., Et al., Rethinking the inception architecture for computer vision, International Conference on Computer Vision and Pattern Recognition, pp. 2818-2826, (2016)
  • [4] Denil M., Shakibi B., Dinh L., Et al., Predicting parameters in deep learning, Advances in Neural Information Processing Systems, pp. 2148-2156, (2013)
  • [5] Srinivas S., Babu R.V., Data-free parameter pruning for deep neural networks
  • [6] Han S., Pool J., Tran J., Et al., Learning both weights and connections for efficient neural network, Advances in Neural Information Processing Systems, pp. 1135-1143, (2015)
  • [7] Mariet Z., Sra S., Diversity networks: neural network compression using determinantal point processes
  • [8] Han S., Mao H., Dally W.J., Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding
  • [9] Simonyan K., Zisserman A., Very deep convolutional networks for large-scale image recognition
  • [10] Iandola F.N., Han S., Moskewicz M.W., Et al., Squeezenet: Alexnet-level accuracy with 50x fewer parameters and &lt
123