Salvaging the Overlooked: Leveraging Class-Aware Contrastive Learning for Multi-Class Anomaly Detection
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913946389184512 |
|---|---|
| author | Fan, Lei Huang, Junjie Di, Donglin Su, Anyang Song, Tianyou Pagnucco, Maurice Song, Yang |
| author_facet | Fan, Lei Huang, Junjie Di, Donglin Su, Anyang Song, Tianyou Pagnucco, Maurice Song, Yang |
| contents | For anomaly detection (AD), early approaches often train separate models for individual classes, yielding high performance but posing challenges in scalability and resource management. Recent efforts have shifted toward training a single model capable of handling multiple classes. However, directly extending early AD methods to multi-class settings often results in degraded performance. In this paper, we investigate this performance degradation observed in reconstruction-based methods, identifying the key issue: inter-class confusion. This confusion emerges when a model trained in multi-class scenarios incorrectly reconstructs samples from one class as those of another, thereby exacerbating reconstruction errors. To this end, we propose a simple yet effective modification, called class-aware contrastive learning (CCL). By explicitly leveraging raw object category information (\eg carpet or wood) as supervised signals, we introduce local CL to refine multiscale dense features, and global CL to obtain more compact feature representations of normal patterns, thereby effectively adapting the models to multi-class settings. Experiments across five datasets validate the effectiveness of our approach, demonstrating significant improvements and superior performance compared to state-of-the-art methods. Notably, ablation studies indicate that pseudo-class labels can achieve comparable performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_04769 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Salvaging the Overlooked: Leveraging Class-Aware Contrastive Learning for Multi-Class Anomaly Detection Fan, Lei Huang, Junjie Di, Donglin Su, Anyang Song, Tianyou Pagnucco, Maurice Song, Yang Computer Vision and Pattern Recognition For anomaly detection (AD), early approaches often train separate models for individual classes, yielding high performance but posing challenges in scalability and resource management. Recent efforts have shifted toward training a single model capable of handling multiple classes. However, directly extending early AD methods to multi-class settings often results in degraded performance. In this paper, we investigate this performance degradation observed in reconstruction-based methods, identifying the key issue: inter-class confusion. This confusion emerges when a model trained in multi-class scenarios incorrectly reconstructs samples from one class as those of another, thereby exacerbating reconstruction errors. To this end, we propose a simple yet effective modification, called class-aware contrastive learning (CCL). By explicitly leveraging raw object category information (\eg carpet or wood) as supervised signals, we introduce local CL to refine multiscale dense features, and global CL to obtain more compact feature representations of normal patterns, thereby effectively adapting the models to multi-class settings. Experiments across five datasets validate the effectiveness of our approach, demonstrating significant improvements and superior performance compared to state-of-the-art methods. Notably, ablation studies indicate that pseudo-class labels can achieve comparable performance. |
| title | Salvaging the Overlooked: Leveraging Class-Aware Contrastive Learning for Multi-Class Anomaly Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.04769 |