文献类型：期刊文献

英文题名：Geometric imbalanced deep learning with feature scaling and boundary sample mining

作者：Wang, Zhe[1,2];Dong, Qida[1,2];Guo, Wei[1,2];Li, Dongdong[1,2];Zhang, Jing[1,2];Du, Wenli[1]

机构：[1]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[2]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China

年份：2022

卷号：126

外文期刊名：PATTERN RECOGNITION

收录：;EI(收录号：20220611594065);WOS:【SCI-EXPANDED(收录号:WOS:000761086100008)】；

基金：This work is supported by Shanghai Science and Technology Program "Distributed and generative few-shot algorithm and the-ory research" under Grant No.20511100600, Natural Science Foun-dation of China under Grant No. 62076094 , Shanghai Science and Technology Program "Federated based cross-domain and cross-task incremental learning" under Grant No. 2151110 080 0, and National Science Foundation of China for Distinguished Young Scholars un-der Grant No. 61725301 .

语种：英文

外文关键词：Imbalance problem; Image classification; Geometric information; Boundary samples mining; Feature scaling

摘要：Data imbalance is a significant factor affecting classification performance in computer vision. In particular, data imbalance is harmful to classification learning and representation learning. To address this issue, this paper proposes a geometric deep learning framework combined with Feature Scaling Module (FSM) and Boundary Samples Mining Module (BSMM). Considering the geometric information in sample distributions of training samples, FSM is proposed to scale the features by hypersphere radius of each class, which improves the representation ability of minority classes. Meanwhile, it is noteworthy that the relationships and information between samples are essential for classification. Therefore, BSMM is proposed to mine the boundary samples by Gabriel Graph that takes the relationships into account. Finally, a loss scheduler is designed to adjust the training process of these two modules. With the scheduler, the model first learns representation and then focuses more on minority classes gradually. Extensive experiments on three benchmark datasets demonstrate the advantages of the proposed learning framework over the state-of-the-art models for solving the imbalance problem. (c) 2022 Elsevier Ltd. All rights reserved.