Multi-View Industrial Anomaly Detection with Epipolar Constrained Cross-View Fusion
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916652522668032 |
|---|---|
| author | Liu, Yifan Xu, Xun Li, Shijie Liao, Jingyi Yang, Xulei |
| author_facet | Liu, Yifan Xu, Xun Li, Shijie Liao, Jingyi Yang, Xulei |
| contents | Multi-camera systems provide richer contextual information for industrial anomaly detection. However, traditional methods process each view independently, disregarding the complementary information across viewpoints. Existing multi-view anomaly detection approaches typically employ data-driven cross-view attention for feature fusion but fail to leverage the unique geometric properties of multi-camera setups. In this work, we introduce an epipolar geometry-constrained attention module to guide cross-view fusion, ensuring more effective information aggregation. To further enhance the potential of cross-view attention, we propose a pretraining strategy inspired by memory bank-based anomaly detection. This approach encourages normal feature representations to form multiple local clusters and incorporate multi-view aware negative sample synthesis to regularize pretraining. We demonstrate that our epipolar guided multi-view anomaly detection framework outperforms existing methods on the state-of-the-art multi-view anomaly detection dataset. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_11088 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multi-View Industrial Anomaly Detection with Epipolar Constrained Cross-View Fusion Liu, Yifan Xu, Xun Li, Shijie Liao, Jingyi Yang, Xulei Computer Vision and Pattern Recognition Multi-camera systems provide richer contextual information for industrial anomaly detection. However, traditional methods process each view independently, disregarding the complementary information across viewpoints. Existing multi-view anomaly detection approaches typically employ data-driven cross-view attention for feature fusion but fail to leverage the unique geometric properties of multi-camera setups. In this work, we introduce an epipolar geometry-constrained attention module to guide cross-view fusion, ensuring more effective information aggregation. To further enhance the potential of cross-view attention, we propose a pretraining strategy inspired by memory bank-based anomaly detection. This approach encourages normal feature representations to form multiple local clusters and incorporate multi-view aware negative sample synthesis to regularize pretraining. We demonstrate that our epipolar guided multi-view anomaly detection framework outperforms existing methods on the state-of-the-art multi-view anomaly detection dataset. |
| title | Multi-View Industrial Anomaly Detection with Epipolar Constrained Cross-View Fusion |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.11088 |