AnomalyDINO-S (1-shot)

Reported on 6 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology6 results

Anomaly DetectiononMVTec AD
Detection AUROC· 2024-05-23
96.6
best: 99.9 (GLASS)
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529
Anomaly DetectiononMVTec AD
Segmentation AUPRO· 2024-05-23
92.7
best: 98.4 (WeakREST-Block)
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529
Anomaly DetectiononMVTec AD
Segmentation AUROC· 2024-05-23
96.8
best: 99.7 (WeakREST-Block)
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529
Anomaly DetectiononVisA
Detection AUROC· 2024-05-23
87.4
best: 99.8 (UniNet)
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529
Anomaly DetectiononVisA
Segmentation AUPRO (until 30% FPR)· 2024-05-23
92.5
best: 96.1 (AnomalyDINO-S (full-shot))
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529
Anomaly DetectiononVisA
Segmentation AUROC· 2024-05-23
97.8
best: 99.1 (Dinomaly ViT-L (model-unified multi-class))
AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2 arXiv:2405.14529