Efficient Anomaly Detection Using Self-Supervised Multi-Cue Tasks

被引:11
|
作者
Jezequel, Loic [1 ,2 ]
Vu, Ngoc-Son [1 ]
Beaudet, Jean [2 ]
Histace, Aymeric [1 ]
机构
[1] CY Cergy Paris Univ, ENSEA, CNRS, ETIS UMR 8051, F-95000 Paris, France
[2] Idemia Ident & Secur, F-95520 Osny, France
关键词
Task analysis; Anomaly detection; Training; Feature extraction; Self-supervised learning; Faces; Neural networks; fine grained classification; self-supervised learning; multi-task learning; one-class learning;
D O I
10.1109/TIP.2022.3231532
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Anomaly detection is important in many real-life applications. Recently, self-supervised learning has greatly helped deep anomaly detection by recognizing several geometric transformations. However these methods lack finer features, usually highly depend on the anomaly type, and do not perform well on fine-grained problems. To address these issues, we first introduce in this work three novel and efficient discriminative and generative tasks which have complementary strength: (i) a piece-wise jigsaw puzzle task focuses on structure cues; (ii) a tint rotation recognition is used within each piece, taking into account the colorimetry information; (iii) and a partial re-colorization task considers the image texture. In order to make the re-colorization task more object-oriented than background-oriented, we propose to include the contextual color information of the image border via an attention mechanism. We then present a new out-of-distribution detection function and highlight its better stability compared to existing methods. Along with it, we also experiment different score fusion functions. Finally, we evaluate our method on an extensive protocol composed of various anomaly types, from object anomalies, style anomalies with fine-grained classification to local anomalies with face anti-spoofing datasets. Our model significantly outperforms state-of-the-art with up to 36% relative error improvement on object anomalies and 40% on face anti-spoofing problems.
引用
收藏
页码:807 / 821
页数:15
相关论文
共 50 条
  • [1] Hyperspectral anomaly detection with self-supervised anomaly prior
    Liu, Yidan
    Jiang, Kai
    Xie, Weiying
    Zhang, Jiaqing
    Li, Yunsong
    Fang, Leyuan
    NEURAL NETWORKS, 2025, 187
  • [2] Self-Supervised Anomaly Detection With Neural Transformations
    Qiu, Chen
    Kloft, Marius
    Mandt, Stephan
    Rudolph, Maja
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2025, 47 (03) : 2170 - 2185
  • [3] SELF-SUPERVISED ANOMALY DETECTION FOR NARROWBAND SETI
    Zhang, Yunfan Gerry
    Won, Ki Hyun
    Son, Seung Woo
    Siemion, Andrew
    Croft, Steve
    2018 IEEE GLOBAL CONFERENCE ON SIGNAL AND INFORMATION PROCESSING (GLOBALSIP 2018), 2018, : 1114 - 1118
  • [4] Self-supervised anomaly detection for new physics
    Dillon, Barry M.
    Mastandrea, Radha
    Nachman, Benjamin
    PHYSICAL REVIEW D, 2022, 106 (05)
  • [5] A self-supervised anomaly detection algorithm with interpretability
    Wu, Zhichao
    Yang, Xin
    Wei, Xiaopeng
    Yuan, Peijun
    Zhang, Yuanping
    Bai, Jianming
    EXPERT SYSTEMS WITH APPLICATIONS, 2024, 237
  • [6] Self-Supervised Autoencoders for Visual Anomaly Detection
    Bauer, Alexander
    Nakajima, Shinichi
    Mueller, Klaus-Robert
    MATHEMATICS, 2024, 12 (24)
  • [7] Anomaly Detection on Electroencephalography with Self-supervised Learning
    Xu, Junjie
    Zheng, Yaojia
    Mao, Yifan
    Wang, Ruixuan
    Zheng, Wei-Shi
    2020 IEEE INTERNATIONAL CONFERENCE ON BIOINFORMATICS AND BIOMEDICINE, 2020, : 363 - 368
  • [8] Multi-cue Contingency Detection
    Jinhan Lee
    Crystal Chao
    Aaron F. Bobick
    Andrea L. Thomaz
    International Journal of Social Robotics, 2012, 4 : 147 - 161
  • [9] Multi-cue Contingency Detection
    Lee, Jinhan
    Chao, Crystal
    Bobick, Aaron F.
    Thomaz, Andrea L.
    INTERNATIONAL JOURNAL OF SOCIAL ROBOTICS, 2012, 4 (02) : 147 - 161
  • [10] Efficient time series anomaly detection by multiresolution self-supervised discriminative network
    Huang, Desen
    Shen, Lifeng
    Yu, Zhongzhong
    Zheng, Zhenjing
    Huang, Min
    Ma, Qianli
    Neurocomputing, 2022, 491 : 261 - 272