Quasibinary Classifier for Images with Zero and Multiple Labels

被引:0
作者
Liao, Shuai [1 ]
Gavves, Efstratios [1 ]
Oh, ChangYong [1 ]
Snoek, Cees G. M. [1 ]
机构
[1] Univ Amsterdam, Amsterdam, Netherlands
来源
2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR) | 2021年
关键词
Binary classifier; softmax classifier; image classification;
D O I
10.1109/ICPR48806.2021.9412933
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The softmax and binary classifier are commonly preferred for image classification applications. However, as softmax is specifically designed for categorical classification, it assumes each image has just one class label. This limits its applicability for problems where the number of labels does not equal one, most notably zero- and multi-label problems. In these challenging settings, binary classifiers are, in theory, better suited. However, as they ignore the correlation between classes, they are not as accurate and scalable in practice. In this paper, we start from the observation that the only difference between binary and softmax classifiers is their normalization function. Specifically, while the binary classifier self-normalizes its score, the softmax classifier combines the scores from all classes before normalisation. On the basis of this observation we introduce a normalization function that is learnable, constant, and shared between classes and data points. By doing so, we arrive at a new type of binary classifier that we coin quasibinary classifier. We show in a variety of image classification settings, and on several datasets, that quasibinary classifiers are considerably better in classification settings where regular binary and softmax classifiers suffer, including zerolabel and multi-label classification. What is more, we show that quasibinary classifiers yield well-calibrated probabilities allowing for direct and reliable comparisons, not only between classes but also between data points.
引用
收藏
页码:8743 / 8750
页数:8
相关论文
共 35 条
  • [1] [Anonymous], 2018, ECCV, DOI DOI 10.1007/978-3-030-01216-8_12
  • [2] Chen H., 2018, MIDL
  • [3] Chua T.-S., 2009, P ACM INT C IM VID R, P1
  • [4] Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848
  • [5] Devlin J., 2014, ACL
  • [6] Fei Wu, 2015, IEEE Transactions on Big Data, V1, P109, DOI 10.1109/TBDATA.2015.2497270
  • [7] Glorot X., 2010, Proceedings of the thirteenth international conference on artificial intelligence and statistics, P249, DOI DOI 10.1109/LGRS.2016.2565705
  • [8] Gong Y., 2013, CoRR
  • [9] Guo CA, 2017, PR MACH LEARN RES, V70
  • [10] Identity Mappings in Deep Residual Networks
    He, Kaiming
    Zhang, Xiangyu
    Ren, Shaoqing
    Sun, Jian
    [J]. COMPUTER VISION - ECCV 2016, PT IV, 2016, 9908 : 630 - 645