Anomaly Segmentation on VisA (AUC-P, AP-P, F1-P)
99.2AUC-PDinomaly2
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Dinomaly2Learning Paradigm=Unsupervised2026.04 | 99.2 | 54.7 | 57.1 | |
| INP-FormerLearning Paradigm=Unsupervised2026.04 | 98.9 | 51.2 | 54.7 | |
| MambaADLearning Paradigm=Unsupervised2026.04 | 98.5 | 39.4 | 44 | |
| GLASSLearning Paradigm=Synthesize-based2026.04 | 98.5 | 45.6 | 48.4 | |
| PatchCoreLearning Paradigm=Unsupervised2026.04 | 98.4 | 48.6 | 49.7 | |
| ViTADLearning Paradigm=Unsupervised2026.04 | 98.2 | 36.6 | 41.1 | |
| GroundingAnomalyBackbone=U-Net, Learning Paradigm=Synthesize-based2026.04 | 98.2 | 67.2 | 63.5 | |
| FAPromptPrompt Category=Prompt Learning, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 95.7 | 21.4 | — | |
| CoPSPrompt Category=Prompt Learning, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 95.7 | 23.4 | — | |
| AnomalyCLIPPrompt Category=Prompt Learning, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 95.5 | 21.3 | — | |
| AdaCLIPPrompt Category=Prompt Design, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 95.1 | 29.2 | — | |
| CLIP-ADPrompt Category=Prompt Design, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 95 | 26.3 | — | |
| APRIL-GANPrompt Category=Prompt Design, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 94.2 | 25.7 | — | |
| DRÆMLearning Paradigm=Synthesize-based2026.04 | 92.9 | 17.2 | 23 | |
| WinCLIPPrompt Category=Prompt Design, Setting=Zero-Shot Anomaly Detection (ZSAD)2025.08 | 79.6 | 5 | — |