文献类型：期刊文献

英文题名：Learning Semantically Enhanced Feature for Fine-Grained Image Classification

作者：Luo, Wei[1];Zhang, Hengmin[2];Li, Jun[3];Wei, Xiu-Shen[3]

机构：[1]South China Agr Univ, Guangzhou 510000, Peoples R China;[2]East China Univ Sci & Technol, Shanghai 200237, Peoples R China;[3]Nanjing Univ Sci & Technol, Nanjing 210094, Peoples R China

年份：2020

卷号：27

起止页码：1545

外文期刊名：IEEE SIGNAL PROCESSING LETTERS

收录：;EI(收录号：20212010348879);WOS:【SCI-EXPANDED(收录号:WOS:000569564700004)】；

基金：This work was supported in part by the NSFC under Grant 61702197 and Grant 61906067 and in part by the NSFGD under Grant 2020A151501813 andGrant 2017A030310261. The associate editor coordinating the reviewof this manuscript and approving it for publication was Prof. Dezhong Peng. (Wei Luo and Hengmin Zhang contributed equally to this work.) (Corresponding author: Wei Luo.)

语种：英文

外文关键词：Semantics; Training; Birds; Feature extraction; Correlation; Entropy; Dogs; Image classification; visual categorization; feature learning

摘要：We aim to provide a computationally cheap yet effective approach for fine-grained image classification (FGIC) in this letter. Unlike previous methods that rely on complex part localization modules, our approach learns fine-grained features by enhancing the semantics of sub-features of a global feature. Specifically, we first achieve the sub-feature semantic by arranging feature channels of a CNN into different groups through channel permutation. Meanwhile, to enhance the discriminability of sub-features, the groups are guided to be activated on object parts with strong discriminability by a weighted combination regularization. Our approach is parameter parsimonious and can be easily integrated into the backbone model as a plug-and-play module for end-to-end training with only image-level supervision. Experiments verified the effectiveness of our approach and validated its comparable performance to the state-of-the-art methods. Code is available at https://github.com/cswluo/SEF.