TR2025-077

TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection


    •  Jung, Y.G., Park, J., Yoon, J., Peng, K.-C., Kim, W., Teoh, A.B.J., Camps, O., "TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection", IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2025.
      BibTeX TR2025-077 PDF Presentation
      • @inproceedings{Jung2025jun,
      • author = {{{Jung, Yoon G. and Park, Jaewoo and Yoon, Jaeho and Peng, Kuan-Chuan and Kim, Wonchul and Teoh, Andrew B. J. and Camps, Octavia}}},
      • title = {{{TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection}}},
      • booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
      • year = 2025,
      • month = jun,
      • url = {https://www.merl.com/publications/TR2025-077}
      • }
  • MERL Contact:
  • Research Areas:

    Artificial Intelligence, Computer Vision, Machine Learning

Abstract:

We aim to solve unsupervised anomaly detection in a practical challenging environment where the normal dataset is both contaminated with defective regions and its product class distribution is tailed but unknown. We observe that existing models suffer from tail-versus-noise trade-off where if a model is robust against pixel noise, then its performance deteriorates on tail class samples, and vice versa. To mitigate the issue, we handle the tail class and noise samples independently. To this end, we propose TailSampler, a novel class size predictor that estimates the class cardinality of samples based on a symmetric assumption on the class-wise distribution of embedding similarities. TailSampler can be utilized to sample the tail class samples exclusively, allowing to handle them separately. Based on these facets, we build a memory-based anomaly detection model TailedCore, whose memory both well captures tail class information and is noise-robust. We extensively validate the effectiveness of TailedCore on the unsupervised long-tail noisy anomaly detection setting, and show that TailedCore outper- forms the state-of-the-art in most settings. Code is available in TailedCore.

 

  • Related Publication

  •  Jung, Y.G., Park, J., Yoon, J., Peng, K.-C., Kim, W., Teoh, A.B.J., Camps, O., "TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection", arXiv, April 2025.
    BibTeX arXiv
    • @article{Jung2025apr,
    • author = {Jung, Yoon G. and Park, Jaewoo and Yoon, Jaeho and Peng, Kuan-Chuan and Kim, Wonchul and Teoh, Andrew B. J. and Camps, Octavia},
    • title = {{TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection}},
    • journal = {arXiv},
    • year = 2025,
    • month = apr,
    • url = {https://arxiv.org/abs/2504.02775}
    • }