Accelerating deep neural network filter pruning with mask-aware convolutional computations on modern CPUs

doi:10.1016/j.neucom.2022.07.006

	Accelerating deep neural network filter pruning with mask-aware convolutional computations on modern CPUs
	Ma, Xiu 1,2; Li, Guangli 3,4; Liu, Lei 1,2; Liu, Huaxiao 1,2; Wang, Xueying 3,4
	2022-09-21
发表期刊	NEUROCOMPUTING
ISSN	0925-2312
卷号	505 页码:375-387
摘要	Filter pruning, a representative model compression technique, has been widely used to compress and accelerate sophisticated deep neural networks on resource-constrained platforms. Nevertheless, most studies focus on reducing the cost of model inference, whereas the heavy burden of the pruning optimiza-tion process is neglected. In this paper, we propose MaskACC, a mask-aware convolutional computation method, which accelerates the prevailing mask-based filter pruning process on modern CPU platforms. MaskACC dynamically reorganizes the tensors used in convolutions with the mask information to avoid unnecessary computations, thereby improving the computational efficiency of the pruning process. Evaluation with state-of-the-art neural network models on CPU cloud platforms demonstrates the effec-tiveness of our method, which achieves up to 1.61x speedup under commonly-used pruning rates, com-pared to conventional computations. (c) 2022 Elsevier B.V. All rights reserved.
关键词	Deep learning systems Neural network compression Filter pruning
DOI	10.1016/j.neucom.2022.07.006
收录类别	SCI
语种	英语
资助项目	National Key R&D Program of China[2021ZD0110101] ; National Natural Science Foundation of China[61872043] ; CCF- Huawei Populus Grove Fund ; Fundamental Research Funds for the Central Universities
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:000861364900010
出版者	ELSEVIER
引用统计	被引频次：4[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/19807
专题	中国科学院计算技术研究所期刊论文
通讯作者	Li, Guangli
作者单位	1.Jilin Univ, Coll Comp Sci & Technol, Changchun, Peoples R China 2.Jilin Univ, Key Lab Symbol Computat & Knowledge Engn, Minist Educ, Changchun, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, State Key Lab Processors, Beijing, Peoples R China 4.Univ Chinese Acad Sci, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Ma, Xiu,Li, Guangli,Liu, Lei,et al. Accelerating deep neural network filter pruning with mask-aware convolutional computations on modern CPUs[J]. NEUROCOMPUTING,2022,505:375-387.
APA	Ma, Xiu,Li, Guangli,Liu, Lei,Liu, Huaxiao,&Wang, Xueying.(2022).Accelerating deep neural network filter pruning with mask-aware convolutional computations on modern CPUs.NEUROCOMPUTING,505,375-387.
MLA	Ma, Xiu,et al."Accelerating deep neural network filter pruning with mask-aware convolutional computations on modern CPUs".NEUROCOMPUTING 505(2022):375-387.