Prototype-based classifier learning for long-tailed visual recognition

作     者：Xiu-Shen WEI Shu-Lin XU Hao CHEN Liang XIAO Yuxin PENG Xiu-Shen WEI;Shu-Lin XU;Hao CHEN;Liang XIAO;Yuxin PENG

作者机构：School of Computer Science and Engineering Nanjing University of Science and Technology Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Educationand Jiangsu Key Lab of Image and Video Understanding for Social Security Wangxuan Institute of Computer Technology Peking University State Key Laboratory of Integrated Services Networks Xidian University 

出 版 物：《Science China(Information Sciences)》 (中国科学:信息科学(英文版))

年 卷 期：2022年第65卷第6期

页      面：62-76页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 080203[工学-机械设计及理论] 0835[工学-软件工程] 0802[工学-机械工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by National Key R&D Program of China (Grant No. 2021YFA1001100) National Natural Science Foundation of China (Grant Nos. 61925201, 62132001, U21B2025, 61871226) Natural Science Foundation of Jiangsu Province of China (Grant No. BK20210340) Fundamental Research Funds for the Central Universities (Grant No.30920041111) CAAI-Huawei MindSpore Open Fund, and Beijing Academy of Artificial Intelligence (BAAI) 

主　　题：long-tailed distribution categorical prototype classifier generation classifier calibration class imbalance 

摘      要：In this paper, we tackle the long-tailed visual recognition problem from the categorical prototype perspective by proposing a prototype-based classifier learning(PCL) method. Specifically, thanks to the generalization ability and robustness, categorical prototypes reveal their advantages of representing the category semantics. Coupled with their class-balance characteristic, categorical prototypes also show potential for handling data imbalance. In our PCL, we propose to generate the categorical classifiers based on the prototypes by performing a learnable mapping function. To further alleviate the impact of imbalance on classifier generation, two kinds of classifier calibration approaches are designed from both prototype-level and example-level aspects. Extensive experiments on five benchmark datasets, including the large-scale iNaturalist, Places-LT,and ImageNet-LT, justify that the proposed PCL can outperform state-of-the-arts. Furthermore, validation experiments can demonstrate the effectiveness of tailored designs in PCL for long-tailed problems.