模式识别与人工智能
Thursday, Apr. 3, 2025 Home      About Journal      Editorial Board      Instructions      Ethics Statement      Contact Us                   中文
  2021, Vol. 34 Issue (11): 1038-1048    DOI: 10.16451/j.cnki.issn1003-6059.202111007
Deep Learning Design and Application Current Issue| Next Issue| Archive| Adv Search |
Lightweight Model Construction Based on Neural Architecture Search
YAO Xiao1, SHI Yewei1, HUO Guanying1, XU Ning1
1. College of Internet of Things Engineering, Hohai University, Changzhou 213022

Download: PDF (790 KB)   HTML (1 KB) 
Export: BibTeX | EndNote (RIS)      
Abstract  The traditional deep neural network cannot be deployed on the edge devices with limited computing capacity due to numerous parameters and high computation. In this paper, a lightweight network based on neural architecture search is specially designed to solve this problem. Convolution units of different groups are regarded as search space, and neural architecture search is utilized to obtain both the group structure and the overall architecture of the network. In the meanwhile, a cycle annealing search strategy is put forward to solve the multi-objective optimization problem of neural architecture search with the consideration of the accuracy and the computation cost of the model. Experiments on datasets show that the proposed network model achieves a better performance than the state-of-the-art methods.
Key wordsLightweight Network      Model Compression      Group Convolution      Neural Architecture Search      Multi-objective Optimization     
Received: 07 January 2021     
ZTFLH: TP 18  
Fund:National Natural Science Foundation of China(No.61501170、41876097), Fundamental Research Funds for the Central Universities(No.B20020205), Key Research and Development Program of Jiangsu Province(No.BK20192004、BE20
18004-04),Open Research Fund of State Key Laboratory of Bio-electronics of Southeast University(No.2019005)
Corresponding Authors: XU Ning, Ph.D., associate professor. His research interests include pattern recognition.   
About author:: YAO Xiao, Ph.D., associate professor. His
research interests include pattern recognition and computer vision.
SHI Yewei, master student. His research interests include pattern recognition and neural network.
HUO Guanying, Ph.D., professor. His research interests include image processing.
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
YAO Xiao
SHI Yewei
HUO Guanying
XU Ning
Cite this article:   
YAO Xiao,SHI Yewei,HUO Guanying等. Lightweight Model Construction Based on Neural Architecture Search[J]. , 2021, 34(11): 1038-1048.
URL:  
http://manu46.magtech.com.cn/Jweb_prai/EN/10.16451/j.cnki.issn1003-6059.202111007      OR     http://manu46.magtech.com.cn/Jweb_prai/EN/Y2021/V34/I11/1038
Copyright © 2010 Editorial Office of Pattern Recognition and Artificial Intelligence
Address: No.350 Shushanhu Road, Hefei, Anhui Province, P.R. China Tel: 0551-65591176 Fax:0551-65591176 Email: bjb@iim.ac.cn
Supported by Beijing Magtech  Email:support@magtech.com.cn