A lightweight convolutional network based on pruning algorithm for YOLO

Guanyu Liu, Yuzhao Li, Yuanchen Song, Yumeng Liu, Xiaofeng Xu, Zhen Zhao, Ruiheng Zhang*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

With the rapid development of deep learning, neural network models have become increasingly complicated, leading to larger storage space requirements and slower reasoning speed. These factors make it difficult to be deployed on resourcelimited platforms. To alleviate this problem, network pruning, an effective model compression method, is commonly performed in a deep neural network. However, traditional pruning methods simply set redundant weights to zero, thus failing to achieve the acceleration effect. In this paper, a channel-wise model scaling method is proposed to reduce the model size and speed up reasoning by structurally removing the redundant filters in convolutional layers. To make the residual block more sparse, we develop a pruning method for residual cells. Experimental results on the YOLOv3 detector show that our proposed approach achieves a 70.6% parameter compression ratio without compromising accuracy.

Original languageEnglish
Title of host publicationFourteenth International Conference on Graphics and Image Processing, ICGIP 2022
EditorsLiang Xiao, Jianru Xue
PublisherSPIE
ISBN (Electronic)9781510666313
DOIs
Publication statusPublished - 2023
Event14th International Conference on Graphics and Image Processing, ICGIP 2022 - Nanjing, China
Duration: 21 Oct 202223 Oct 2022

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume12705
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

Conference14th International Conference on Graphics and Image Processing, ICGIP 2022
Country/TerritoryChina
CityNanjing
Period21/10/2223/10/22

Keywords

  • deep learning
  • model compression
  • network pruning

Fingerprint

Dive into the research topics of 'A lightweight convolutional network based on pruning algorithm for YOLO'. Together they form a unique fingerprint.

Cite this