Anomaly Detection on VAD high-shot supervised
96.5AUROC (Classification)AllAD + WRN + SegAD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AllAD + WRN + SegADBase Method=PatchCore, FastFlow, RD4AD, and EfficientAD, Classifier=Wide ResNet502024.05 | 96.5 | 16.4 | |
| WRNCategory=Supervised Classifier, Backbone=Wide ResNet502024.05 | 95 | 32.8 | |
| PatchCore + SegADBase Method=PatchCore2024.05 | 91.7 | 48 | |
| EfficientAD + SegADBase Method=EfficientAD2024.05 | 91.4 | 36.8 | |
| RD4AD + SegADBase Method=RD4AD2024.05 | 90.2 | 46.9 | |
| DRACategory=Supervised Anomaly Detection2024.05 | 87.4 | 60.9 | |
| DevNetCategory=Supervised Anomaly Detection2024.05 | 86.9 | 66.6 | |
| FastFlow + SegADBase Method=FastFlow2024.05 | 86.9 | 56.9 |