AnomalyCLIP

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· 2023-10-29
91.5
best: 99.9 (GLASS)
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961
Anomaly DetectiononMVTec AD
Segmentation AUPRO· 2023-10-29
81.4
best: 98.4 (WeakREST-Block)
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961
Anomaly DetectiononMVTec AD
Segmentation AUROC· 2023-10-29
91.1
best: 99.7 (WeakREST-Block)
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961
Anomaly DetectiononVisA
Detection AUROC· uses extra data· 2023-10-29
82.1
best: 99.8 (UniNet)
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961
Anomaly DetectiononVisA
Segmentation AUPRO· uses extra data· 2023-10-29
87
best: 96 (DiffusionAD)
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961
Anomaly DetectiononVisA
Segmentation AUROC· uses extra data· 2023-10-29
95.5
best: 99.1 (Dinomaly ViT-L (model-unified multi-class))
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection arXiv:2310.18961