Acknowledging the Unknown for Multi-label Learning with Single Positive Labels

被引:4
|
作者
Zhou, Donghao [1 ,2 ]
Chen, Pengfei [3 ]
Wang, Qiong [1 ]
Chen, Guangyong [4 ]
Heng, Pheng-Ann [1 ,5 ]
机构
[1] Chinese Acad Sci, Shenzhen Inst Adv Technol, Guangdong Prov Key Lab Comp Vis & Virtual Real Te, Shenzhen, Peoples R China
[2] Univ Chinese Acad Sci, Beijing, Peoples R China
[3] Tencent Technol, Shenzhen, Peoples R China
[4] Zhejiang Lab, Hangzhou, Peoples R China
[5] Chinese Univ Hong Kong, Hong Kong, Peoples R China
来源
基金
中国国家自然科学基金; 国家重点研发计划;
关键词
Weakly supervised learning; Single positive multi-label learning; Entropy maximization; Pseudo-labeling;
D O I
10.1007/978-3-031-20053-3_25
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Due to the difficulty of collecting exhaustive multi-label annotations, multi-label datasets often contain partial labels. We consider an extreme of this weakly supervised learning problem, called single positive multi-label learning (SPML), where each multi-label training image has only one positive label. Traditionally, all unannotated labels are assumed as negative labels in SPML, which introduces false negative labels and causes model training to be dominated by assumed negative labels. In this work, we choose to treat all unannotated labels from an alternative perspective, i.e. acknowledging they are unknown. Hence, we propose entropy-maximization (EM) loss to attain a special gradient regime for providing proper supervision signals. Moreover, we propose asymmetric pseudo-labeling (APL), which adopts asymmetric-tolerance strategies and a self-paced procedure, to cooperate with EM loss and then provide more precise supervision. Experiments show that our method significantly improves performance and achieves state-of-the-art results on all four benchmarks. Code is available at https://github.com/Correr-Zhou/SPML-AckTheUnknown.
引用
收藏
页码:423 / 440
页数:18
相关论文
共 50 条
  • [41] Learning Common and Label-Specific Features for Multi-Label Classification With Missing Labels
    Li, Runxin
    Ouyang, Zexian
    Shang, Zhenhong
    Jia, Lianyin
    Li, Xiaowu
    [J]. IEEE ACCESS, 2024, 12 : 81170 - 81195
  • [42] Improving multi-label classification with missing labels by learning label-specific features
    Huang, Jun
    Qin, Feng
    Zheng, Xiao
    Cheng, Zekai
    Yuan, Zhixiang
    Zhang, Weigang
    Huang, Qingming
    [J]. INFORMATION SCIENCES, 2019, 492 : 124 - 146
  • [43] Learning Low-Rank Label Correlations for Multi-label Classification with Missing Labels
    Xu, Linli
    Wang, Zhen
    Shen, Zefan
    Wang, Yubo
    Chen, Enhong
    [J]. 2014 IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2014, : 1067 - 1072
  • [44] Bayesian Multi-label Learning with Sparse Features and Labels, and Label Co-occurrences
    Zhao, He
    Raiy, Piyush
    Du, Lan
    Buntine, Wray
    [J]. INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 84, 2018, 84
  • [45] Robust Recurrent Classifier Chains for Multi-Label Learning with Missing Labels
    Gerych, Walter
    Hartvigsen, Thomas
    Buquicchio, Luke
    Agu, Emmanuel
    Rundensteiner, Elke
    [J]. PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2022, 2022, : 582 - 591
  • [46] Multi-label learning with incomplete labels via dual manifold mappings
    Huang, Rui
    Xu, Zhilei
    [J]. INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS, 2024, 15 (06) : 2495 - 2511
  • [47] Multi-Label Learning with Missing Labels via Common and Label-Specific Features
    Sun, Mengxuan
    Li, Peipei
    Li, Junlong
    Hu, Xuegang
    [J]. 2022 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2022,
  • [48] Improving Multi-label Learning with Missing Labels by Structured Semantic Correlations
    Yang, Hao
    Zhou, Joey Tianyi
    Cai, Jianfei
    [J]. COMPUTER VISION - ECCV 2016, PT I, 2016, 9905 : 835 - 851
  • [49] Multi-label Learning with Missing Labels Using Mixed Dependency Graphs
    Baoyuan Wu
    Fan Jia
    Wei Liu
    Bernard Ghanem
    Siwei Lyu
    [J]. International Journal of Computer Vision, 2018, 126 : 875 - 896
  • [50] Multi-label Learning with Missing Labels Using Mixed Dependency Graphs
    Wu, Baoyuan
    Jia, Fan
    Liu, Wei
    Ghanem, Bernard
    Lyu, Siwei
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2018, 126 (08) : 875 - 896