TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913161945284608 |
|---|---|
| author | Jung, Yoon Gyo Park, Jaewoo Yoon, Jaeho Peng, Kuan-Chuan Kim, Wonchul Teoh, Andrew Beng Jin Camps, Octavia |
| author_facet | Jung, Yoon Gyo Park, Jaewoo Yoon, Jaeho Peng, Kuan-Chuan Kim, Wonchul Teoh, Andrew Beng Jin Camps, Octavia |
| contents | 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 outperforms the state-of-the-art in most settings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_02775 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection Jung, Yoon Gyo Park, Jaewoo Yoon, Jaeho Peng, Kuan-Chuan Kim, Wonchul Teoh, Andrew Beng Jin Camps, Octavia Computer Vision and Pattern Recognition Machine Learning 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 outperforms the state-of-the-art in most settings. |
| title | TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2504.02775 |