中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 合肥物质科学研究院
  3. 中国科学院合肥物质科学研究院
AM-ResNet: Low-energy-consumption addition-multiplication hybrid ResNet for pest recognition

文献类型：期刊论文

作者Zhang, Li1,2; Du, Jianming2,3; Dong, Shifeng1,2; Wang, Fenmei1,2; Xie, Chengjun2; Wang, Rujing1,2,3,4
刊名COMPUTERS AND ELECTRONICS IN AGRICULTURE
出版日期2022-11-01
卷号202
关键词Pest recognition Convolutional neural networks Addernet Energy consumption
ISSN号0168-1699
DOI10.1016/j.compag.2022.107357
通讯作者Du, Jianming(djming@iim.ac.cn) ; Wang, Rujing(rjwang@iim.ac.cn)
英文摘要Pest recognition technology has rapidly progressed in a short period with the development of deep convolutional neural networks. However, the vast calculation burden of these networks requires massive energy, especially for specific applications such as all-day-working pest monitoring systems, which process images from dozens of devices uninterruptedly. This paper proposes a low-energy-consumption hybrid ResNet structure - AM-ResNet, consisting of addition-based and multiplication-based convolutional layers to address this problem. This paper presents an optimal AM-ResNet design method through a detailed experimental analysis of the performance differences between building blocks in two typical ResNet variants, ResNet20 and ResNet32. Then, this method is applied to construct a deep AM-ResNet for pest recognition, which significantly reduces the energy consumption at the cost of an acceptable accuracy loss. Experiments show that the proposed network performs well in our proposed pest dataset PEST20. Extensive experiments demonstrate that the network can save more than 40% energy by losing less than 2% accuracy in IP102 and VOC2007. In addition, by analyzing the visualization re-sults, this paper summarizes the advantages of the two convolutions. It presents the application direction for the hybrid networks.
资助项目national natural science foundation of China[31671586] ; Dean's Fund of Hefei Institute of Physical Science, Chinese Academy of Sciences[YZJJ2022QN32]
WOS研究方向Agriculture ; Computer Science
语种英语
WOS记录号WOS:000874972800001
出版者ELSEVIER SCI LTD
资助机构national natural science foundation of China ; Dean's Fund of Hefei Institute of Physical Science, Chinese Academy of Sciences
源URL[http://ir.hfcas.ac.cn:8080/handle/334002/129867]  
专题中国科学院合肥物质科学研究院
通讯作者Du, Jianming; Wang, Rujing
作者单位1.Univ Sci & Technol China, Hefei 230026, Peoples R China
2.Chinese Acad Sci, Inst Intelligent Machines, Hefei Inst Phys Sci, Hefei 230031, Peoples R China
3.Chinese Acad Sci, Inst Intelligent Machines, Hefei Inst Phys Sci, Hefei, Peoples R China
4.Univ Sci & Technol China, Hefei, Peoples R China
推荐引用方式
GB/T 7714		 Zhang, Li,Du, Jianming,Dong, Shifeng,et al. AM-ResNet: Low-energy-consumption addition-multiplication hybrid ResNet for pest recognition[J]. COMPUTERS AND ELECTRONICS IN AGRICULTURE,2022,202.
APA Zhang, Li,Du, Jianming,Dong, Shifeng,Wang, Fenmei,Xie, Chengjun,&Wang, Rujing.(2022).AM-ResNet: Low-energy-consumption addition-multiplication hybrid ResNet for pest recognition.COMPUTERS AND ELECTRONICS IN AGRICULTURE,202.
MLA Zhang, Li,et al."AM-ResNet: Low-energy-consumption addition-multiplication hybrid ResNet for pest recognition".COMPUTERS AND ELECTRONICS IN AGRICULTURE 202(2022).

入库方式： OAI收割

来源：合肥物质科学研究院

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。