Hannah M. Schlüter, Jeremy Tan, Benjamin Hou, Bernhard Kainz
We introduce a simple and intuitive self-supervision task, Natural Synthetic Anomalies (NSA), for training an end-to-end model for anomaly detection and localization using only normal training data. NSA integrates Poisson image editing to seamlessly blend scaled patches of various sizes from separate images. This creates a wide range of synthetic anomalies which are more similar to natural sub-image irregularities than previous data-augmentation strategies for self-supervised anomaly detection. We evaluate the proposed method using natural and medical images. Our experiments with the MVTec AD dataset show that a model trained to localize NSA anomalies generalizes well to detecting real-world a priori unknown types of manufacturing defects. Our method achieves an overall detection AUROC of 97.2 outperforming all previous methods that learn without the use of additional datasets. Code available at https://github.com/hmsch/natural-synthetic-anomalies.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Anomaly Detection | AeBAD-V | Detection AUROC | 64.6 | NSA |
| Anomaly Detection | AeBAD-S | Detection AUROC | 56.7 | NSA |
| Anomaly Detection | AeBAD-S | Segmentation AUPRO | 45.9 | NSA |
| Anomaly Detection | MVTec AD | Detection AUROC | 97.2 | NSA |
| Anomaly Detection | MVTec AD | Segmentation AUPRO | 91 | NSA |
| Anomaly Detection | MVTec AD | Segmentation AUROC | 96.3 | NSA |
| Anomaly Detection | GoodsAD | AUPR | 71.8 | NSA |
| Anomaly Detection | GoodsAD | AUROC | 67.3 | NSA |
| 2D Classification | GoodsAD | AUPR | 71.8 | NSA |
| 2D Classification | GoodsAD | AUROC | 67.3 | NSA |
| Anomaly Classification | GoodsAD | AUPR | 71.8 | NSA |
| Anomaly Classification | GoodsAD | AUROC | 67.3 | NSA |