文献类型：期刊文献

英文题名：Cooperative density-aware representation learning for few-shot visual recognition

作者：Zheng, Zijun;Feng, Xiang[1];Yu, Huiqun;Gao, Mengqi

机构：[1]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China; Shanghai Engn Res Ctr Smart Energy, Shanghai 200237, Peoples R China

年份：2022

卷号：471

起止页码：208

外文期刊名：NEUROCOMPUTING

收录：;EI(收录号：20214811252855);WOS:【SCI-EXPANDED(收录号:WOS:000761907400010)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant Nos. 61772200, 61772201 and 61602175, the Information Development Special Funds of Shanghai Pujiang Talent Program under Grant No. 17PJ1401900, the Information Development Special Funds of Shanghai Economic and Information Commission under Grant No. 201602008, the Open Funds of Shanghai Smart City Collabora-tive Innovation Center.

语种：英文

外文关键词：Few-shot visual recognition; Cooperative density loss; Representation learning; Semantic information

摘要：Few-shot visual recognition has achieved remarkable advances along with the rise of deep learning. Its goal is to learn the model parameter from the base category for transferring it to the novel category with limited annotations. However, most of the existing few-shot visual recognition approaches mainly focus on extracting a global feature representation of the sample, which fails to encode the semantic information. To alleviate this issue, this paper presents a novel cooperative density-aware representation learning approach for few-shot visual recognition. Specifically, we first yield the high-level semantic features of the query set and the support set by leveraging a shared convolutional neural network. A cooperative density loss module is then designed to optimize the model to form the discriminative features by incorporating the density global classification loss and the density few-shot loss. The density few-shot loss conducts the semantic alignment with regional features by the mutual information finding manner while the density global classification loss supervises each regional feature lead to more precise classification. Comprehensive experiments in few-shot visual recognition benchmarks validate the effectiveness and superiority of our proposed approach, and elaborate ablations explain the utility of different modules. (c) 2021 Elsevier B.V. All rights reserved.