Iterative Low-rank Approximation Based on the Redundancy of Each Network Layer

doi:10.1117/12.2589425

IR > 理学院

	Iterative Low-rank Approximation Based on the Redundancy of Each Network Layer
	Yang, Fan1 ; Liu, Weirong1 ; Liu, Jie 2; Liu, Chaorong3 ; Mi, Yanchun 1; Song, Haowen 1
	2021
会议名称	12th International Conference on Graphics and Image Processing (ICGIP)
会议录名称	TWELFTH INTERNATIONAL CONFERENCE ON GRAPHICS AND IMAGE PROCESSING (ICGIP 2020)
卷号	11720
会议日期	NOV 13-15, 2020
会议地点	Xian, PEOPLES R CHINA
出版地	BELLINGHAM
出版者	SPIE-INT SOC OPTICAL ENGINEERING
摘要	Low rank approximation is an effective method in deep neural network (DNN) compression. In view of the fact that the redundancy information content of different network layers is different, a novel iterative low-rank approximation method based on the redundancy of each network layer is proposed. By giving priority to the network layer with higher redundancy, the loss of intrinsic information in each network layer is expected to be reduced and the performance of the compressed model is improved. Experimental results show that the performance of compressed model obtained by this method is improved with a slight reduction in compression ratio. It can be concluded that the proposed method can better retain intrinsic information in the pre-training network.
关键词	Priority compression iterative compression low-rank approximation DNN
DOI	10.1117/12.2589425
收录类别	CPCI-S
语种	英语
WOS研究方向	Computer Science ; Engineering ; Optics ; Imaging Science & Photographic Technology
WOS类目	Computer Science, Software Engineering ; Engineering, Electrical & Electronic ; Optics ; Imaging Science & Photographic Technology
WOS记录号	WOS:000700371000087
ISSN	0277-786X
引用统计	被引频次：1[WOS] [WOS记录] [WOS相关记录]
文献类型	会议论文
条目标识符	https://ir.lut.edu.cn/handle/2XXMBERH/150131
专题	理学院电气工程与信息工程学院党委教师工作部（人事处、教师发展中心）
通讯作者	Yang, Fan
作者单位	1.Lanzhou Univ Technol, Coll Elect & Informat Engn, Lanzhou, Peoples R China; 2.Lanzhou Univ Technol, Natl Demonstrat Ctr Expt Elect & Control Engn Edu, Lanzhou, Peoples R China; 3.Lanzhou Univ Technol, Key Lab Gansu Adv Control Ind Proc, Lanzhou, Peoples R China
第一作者单位	电气工程与信息工程学院
通讯作者单位	电气工程与信息工程学院
推荐引用方式 GB/T 7714	Yang, Fan,Liu, Weirong,Liu, Jie,et al. Iterative Low-rank Approximation Based on the Redundancy of Each Network Layer[C]. BELLINGHAM:SPIE-INT SOC OPTICAL ENGINEERING,2021.