An Approach to Pruning the Structure of Convolutional Neural Networks without Loss of Generalization Ability

被引：0

作者：

Chen, Chaoxiang ^{[1
,2
,3
]}

Kroshchanka, Aliaksandr ^{[4
]}

Golovko, Vladimir ^{[4
,5
]}

Golovko, Olha ^{[4
]}

机构：

[1] Zhejiang Shuren Univ, Sch Informat Sci & Technol, Hangzhou 310015, Peoples R China

[2] Int Sci & Technol Cooperat Base Zhejiang Prov Remo, Hangzhou 310000, Peoples R China

[3] Zhejiang Shuren Univ, Inst Tradit Chinese Med Artificial Intelligence, Hangzhou 310015, Peoples R China

[4] Brest State Tech Univ, Brest 224017, BELARUS

[5] John Paul II Univ Biala Podlaska, PL-21500 Biala Podlaska, Poland

来源：

PATTERN RECOGNITION AND IMAGE ANALYSIS | 2024年 / 34卷 / 02期

关键词：

convolutional neural networks; convolutional restricted Boltzmann machine (CRBM); pruning of neural network parameters; pretraining of convolutional neural networks; computer vision;

D O I：

10.1134/S1054661824700056

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

This paper proposes an approach to pruning the parameters of convolutional neural networks using unsupervised pretraining. The authors demonstrate that the proposed approach makes it possible to reduce the number of configurable parameters of a convolutional neural network without loss of generalization ability. A comparison of the proposed approach and existing pruning techniques is made. The capabilities of the proposed algorithm are demonstrated on classical CIFAR10 and CIFAR100 computer vision samples.

引用

页码：258 / 265

页数：8

共 13 条

[1] Davies R. E., 2022, Advanced Methods and Deep Learning in Computer Vision, DOI [10.1016/c2019-0-03221-9, DOI 10.1016/C2019-0-03221-9]
[2] Golovko V. A., 2015, Lecture Notes in Neuroinformatics, P75
[3] Golovko V, 2015, INT WORKSH INT DATA, P182, DOI 10.1109/IDAACS.2015.7340725
[4] Han S, 2015, ADV NEUR IN, V28
[5] Hinton G, 2015, Arxiv, DOI [arXiv:1503.02531, DOI 10.48550/ARXIV.1503.02531]
[6] A fast learning algorithm for deep belief nets
Hinton, Geoffrey E.
Osindero, Simon
Teh, Yee-Whye
[J]. NEURAL COMPUTATION, 2006, 18 (07) : 1527 - 1554
[7] Hubara I, 2018, J MACH LEARN RES, V18
[8] Iodice G. M., 2022, Tiny ML: Cookbook: Combine Artificial Intelligence and Ultra-Low-Power Embedded Devices to Make the World Smarter
[9] Krizhevsky A., 2009, Technical Report) University of Toronto, Learn. Mult. Layers Features Tiny Images
[10] Method for Reducing Neural-Network Models of Computer Vision
Kroshchanka, A. A.
Golovko, V. A.
Chodyka, M.
[J]. PATTERN RECOGNITION AND IMAGE ANALYSIS, 2022, 32 (02) : 294 - 300

← 1 2 →