Unbiased Subclass Regularization for Semi-Supervised Semantic Segmentation

被引:21
|
作者
Guan, Dayan [1 ]
Huang, Jiaxing [1 ]
Xiao, Aoran [1 ]
Lu, Shijian [1 ]
机构
[1] Nanyang Technol Univ, Singtel Cognit & Artificial Intelligence Lab Ente, Singapore, Singapore
关键词
D O I
10.1109/CVPR52688.2022.00973
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Semi-supervised semantic segmentation learns from small amounts of labelled images and large amounts of unlabelled images, which has witnessed impressive progress with the recent advance of deep neural networks. However, it often suffers from severe class-bias problem while exploring the unlabelled images, largely due to the clear pixel-wise class imbalance in the labelled images. This paper presents an unbiased subclass regularization network (USRN) that alleviates the class imbalance issue by learning class-unbiased segmentation from balanced subclass distributions. We build the balanced subclass distributions by clustering pixels of each original class into multiple subclasses of similar sizes, which provide class-balanced pseudo supervision to regularize the class-biased segmentation. In addition, we design an entropy-based gate mechanism to coordinate learning between the original classes and the clustered subclasses which facilitates subclass regularization effectively by suppressing unconfident subclass predictions. Extensive experiments over multiple public benchmarks show that USRN achieves superior performance as compared with the state-of-the-art.
引用
下载
收藏
页码:9958 / 9968
页数:11
相关论文
共 50 条
  • [1] SEMI-SUPERVISED SEMANTIC SEGMENTATION CONSTRAINED BY CONSISTENCY REGULARIZATION
    Li, Xiaoqiang
    He, Qin
    Dai, Songmin
    Wu, Pin
    Tong, Weiqin
    2020 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO (ICME), 2020,
  • [2] Class Probability Space Regularization for semi-supervised semantic segmentation
    Yin, Jianjian
    Yan, Shuai
    Chen, Tao
    Chen, Yi
    Yao, Yazhou
    COMPUTER VISION AND IMAGE UNDERSTANDING, 2024, 249
  • [3] Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization
    Xu, Hai-Ming
    Liu, Lingqiao
    Bian, Qiuchen
    Yang, Zhen
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [4] Perturbation consistency and mutual information regularization for semi-supervised semantic segmentation
    Wu, Yulin
    Liu, Chang
    Chen, Lei
    Zhao, Dong
    Zheng, Qinghe
    Zhou, Hongchao
    MULTIMEDIA SYSTEMS, 2023, 29 (02) : 511 - 523
  • [5] Perturbation consistency and mutual information regularization for semi-supervised semantic segmentation
    Yulin Wu
    Chang Liu
    Lei Chen
    Dong Zhao
    Qinghe Zheng
    Hongchao Zhou
    Multimedia Systems, 2023, 29 : 511 - 523
  • [6] Transferable Semi-Supervised Semantic Segmentation
    Xiao, Huaxin
    Wei, Yunchao
    Liu, Yu
    Zhang, Maojun
    Feng, Jiashi
    THIRTY-SECOND AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTIETH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / EIGHTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2018, : 7420 - 7427
  • [7] Universal Semi-Supervised Semantic Segmentation
    Kalluri, Tarun
    Varma, Girish
    Chandraker, Manmohan
    Jawahar, C. V.
    2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, : 5258 - 5269
  • [8] Consistency Regularization for Semi-Supervised Semantic Segmentation of Flood Regions from SAR Images
    Savitha, G.
    Girisha, S.
    Sughosh, Pundarika
    Shetty, Dasharathraj K.
    Mymbilly Balakrishnan, Jayaraj
    Paul, Rahul
    Naik, Nithesh
    IEEE Access, 2025, 13 : 9642 - 9653
  • [9] Semantic Segmentation with Active Semi-Supervised Learning
    Rangnekar, Aneesh
    Kanan, Christopher
    Hoffman, Matthew
    2023 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV), 2023, : 5955 - 5966
  • [10] LaserMix for Semi-Supervised LiDAR Semantic Segmentation
    Kong, Lingdong
    Ren, Jiawei
    Pan, Liang
    Liu, Ziwei
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 21705 - 21715