您好，欢迎访问广西农业科学院 机构知识库！

DIEC-ViT: Discriminative information enhanced contrastive vision transformer for the identification of plant diseases in complex environments

文献类型： 外文期刊

作者： Lin, Jianwu 1 ; Chen, Xiaoyulong 4 ; Lou, Lunhong 3 ; You, Lin 3 ; Cernava, Tomislav 6 ; Huang, Dahui 7 ; Qin, Yongbin 1 ; Zhang, Xin 1 ;

作者机构： 1.Guizhou Univ, Coll Comp Sci & Technol, State Key Lab Publ Big Data, Guiyang 550025, Peoples R China

2.Guizhou Univ, Coll Comp Sci & Technol, Text Comp & Cognit Intelligence Engn Res Ctr, Natl Educ Minist, Guiyang 550025, Peoples R China

3.Guizhou Univ, Coll Big Data & Informat Engn, Guiyang 550025, Peoples R China

4.Guizhou Univ, Coll Life Sci, Guiyang 550025, Peoples R China

5.Guizhou Univ, Guizhou Prov Sci & Technol Dept, Guizhou Europe Environm Biotechnol & Agr Informat, Guiyang 550025, Guizhou, Peoples R China

6.Univ Southampton, Fac Environm & Life Sci, Sch Biol Sci, Southampton SO17 1BJ, England

7.Guangxi Acad Agr Sci, Rice Res Inst, Guangxi Key Lab Rice Genet & Breeding, Guangxi Crop Genet Improvement & Biotechnol Lab, Nanning 530007, Peoples R China

关键词： Plant disease recognition; Vision transformers; Discriminative information enhancement; modules; Contrastive learning

期刊名称：EXPERT SYSTEMS WITH APPLICATIONS （ 影响因子：7.5； 五年影响因子：7.8 ）

ISSN： 0957-4174

年卷期： 2025 年 281 卷

页码：

收录情况： SCI

摘要： Recently, vision transformer (ViT)-based methods have made breakthroughs on plant disease recognition tasks and have surpassed convolutional neural network (CNN)-based methods. They are now considered the state-ofthe-art for such methods. However, ViT-based methods usually encode and decode images through global modeling, which introduces a large amount of noise information when dealing with plant disease images in complex environments. In addition, plant disease images in complex environments have significant intra- and inter-class differences, further limiting the performance of ViT-based methods. To address the above limitations, we propose the discriminative information enhanced contrastive vision transformer, in short DIEC-ViT, for plant disease recognition in complex environments. DIEC-ViT contains two key modules, namely, the discriminative information enhancement (DIE) module and the contrastive learning (CL) module. Specifically, the DIE module enhances the perception of discriminative regions of the ViT and suppresses complex backgrounds by counting multi-head self-attention for multi-levels of class tokens. To cope with the problem of intra- and inter-class differences in plant disease images, the CL module is introduced into the ViT to optimize the feature space by reducing the distance between positive pairs and increasing the distance between negative pairs. Extensive experiments verify the effectiveness of the two modules. In addition, DEIC-ViT outperforms state-of-the-art methods with three field plant disease datasets. The obtained results indicate the potential of our approach to drive further development of ViT in the field of plant disease monitoring.

  • 相关文献
作者其他论文 更多>>
置顶
购物车
反馈

© 京ICP备09089781号-29 主办单位：广西农业科学院

学术资源： 中国农业科学院 农业专业知识服务系统 农科联盟资源共建共享平台 中国农业科技文献与信息服务平台 中国农业期刊集成服务平台