Out-Of-Distribution Detection In Unsupervised Continual Learning

被引:3
|
作者
He, Jiangpeng [1 ]
Zhu, Fengqing [1 ]
机构
[1] Purdue Univ, Elmore Family Sch Elect & Comp Engn, W Lafayette, IN 47907 USA
关键词
D O I
10.1109/CVPRW56347.2022.00430
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Unsupervised continual learning aims to learn new tasks incrementally without requiring human annotations. However, most existing methods, especially those targeted on image classification, only work in a simplified scenario by assuming all new data belong to new tasks, which is not realistic if the class labels are not provided. Therefore, to perform unsupervised continual learning in real life applications, an out-of-distribution detector is required at beginning to identify whether each new data corresponds to a new task or already learned tasks, which still remains under-explored yet. In this work, we formulate the problem for Out-of-distribution Detection in Unsupervised Continual Learning (OOD-UCL) with the corresponding evaluation protocol. In addition, we propose a novel OOD detection method by correcting the output bias at first and then enhancing the output confidence for in-distribution data based on task discriminativeness, which can be applied directly without modifying the learning procedures and objectives of continual learning. Our method is evaluated on CIFAR-100 dataset by following the proposed evaluation protocol and we show improved performance compared with existing OOD detection methods under the unsupervised continual learning scenario.
引用
收藏
页码:3849 / 3854
页数:6
相关论文
共 50 条
  • [31] Mitral Regurgitation Recogniton Based on Unsupervised Out-of-Distribution Detection with Residual Diffusion Amplification
    Liu, Zhe
    Zhu, Xiliang
    Han, Tong
    Huang, Yuhao
    Wang, Jian
    Li, Lian
    Wang, Fang
    Ni, Dong
    Gou, Zhongshan
    Yang, Xin
    MACHINE LEARNING IN MEDICAL IMAGING, PT I, MLMI 2024, 2025, 15241 : 52 - 62
  • [32] SS3D: Unsupervised Out-of-Distribution Detection and Localization for Medical Volumes
    Doorenbos, Lars
    Sznitman, Raphael
    Marquez-Neila, Pablo
    BIOMEDICAL IMAGE REGISTRATION, DOMAIN GENERALISATION AND OUT-OF-DISTRIBUTION ANALYSIS, 2022, 13166 : 111 - 118
  • [33] Out-of-Distribution Detection for Automotive Perception
    Nitsch, Julia
    Itkina, Masha
    Senanayake, Ransalu
    Nieto, Juan
    Schmidt, Max
    Siegwart, Roland
    Kochenderfer, Mykel J.
    Cadena, Cesar
    2021 IEEE INTELLIGENT TRANSPORTATION SYSTEMS CONFERENCE (ITSC), 2021, : 2938 - 2943
  • [34] Decoupling MaxLogit for Out-of-Distribution Detection
    Zhang, Zihan
    Xiang, Xiang
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR, 2023, : 3388 - 3397
  • [35] Robust Cough Detection With Out-of-Distribution Detection
    Chen, Yuhan
    Attri, Pankaj
    Barahona, Jeffrey
    Hernandez, Michelle L.
    Carpenter, Delesha
    Bozkurt, Alper
    Lobaton, Edgar
    IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2023, 27 (07) : 3210 - 3221
  • [36] Learning on Graphs with Out-of-Distribution Nodes
    Song, Yu
    Wang, Donglin
    PROCEEDINGS OF THE 28TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, KDD 2022, 2022, : 1635 - 1645
  • [37] Exploring the Limits of Out-of-Distribution Detection
    Fort, Stanislav
    Ren, Jie
    Lakshminarayanan, Balaji
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [38] On Risk Assessment for Out-of-Distribution Detection
    Vasiliuk, Anton
    IEEE ACCESS, 2025, 13 : 18546 - 18568
  • [39] Semantic enhanced for out-of-distribution detection
    Jiang, Weijie
    Yu, Yuanlong
    FRONTIERS IN NEUROROBOTICS, 2022, 16
  • [40] Likelihood Ratios for Out-of-Distribution Detection
    Ren, Jie
    Liu, Peter J.
    Fertig, Emily
    Snoek, Jasper
    Poplin, Ryan
    DePristo, Mark A.
    Dillon, Joshua V.
    Lakshminarayanan, Balaji
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019), 2019, 32