OOD Detection on MVTec Pill (Similar)
93.33AUROCKNN
Evaluation results
| Method | Links | |
|---|---|---|
| KNNBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 93.33 | |
| KNNBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 91.28 | |
| CoPBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 88.27 | |
| DICEBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 87.63 | |
| GradOrthBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 86.86 | |
| ResidualBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 86.09 | |
| ViMBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 85.9 | |
| NMDBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 85.26 | |
| LOFBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 83.78 | |
| ReActBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 83.72 | |
| ViMBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 83.21 | |
| ODINBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 82.69 | |
| CoRPBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 81.92 | |
| KDE (Gaussian)Backbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 81.35 | |
| ASHBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 81.35 | |
| SHEBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 81.09 | |
| WeiPerBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 80.77 | |
| NuSABackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 80.64 | |
| GradNormBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 80.13 | |
| ResidualBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 79.87 | |
| MC-DropoutBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 79.57 | |
| Deep EnsembleBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 79.54 | |
| MCPBackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 78.46 | |
| KNNBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 77.95 | |
| CoPBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 76.68 | |
| XOOD-MBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 75.71 | |
| GRAMBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 74.76 | |
| XOOD-MBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 74.08 | |
| MBMBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 73.18 | |
| CoPBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 72.76 | |
| MBMBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 72.03 | |
| MahalanobisBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 71.86 | |
| LOFBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 71.79 | |
| FeatureNormBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 71.67 | |
| MahalanobisBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 71.15 | |
| NuSABackbone=ViT-B/32, Method Type=Feature-based2026.02 | 70.64 | |
| MahalanobisBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 70.28 | |
| SHEBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 70.26 | |
| MBMBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 70.09 | |
| XOOD-MBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 70.08 | |
| CoRPBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 69.87 | |
| CoRPBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 69.42 | |
| TAPUUDBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 68.9 | |
| PCXBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 68.56 | |
| NACBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 68.27 | |
| NMDBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 67.37 | |
| Deep EnsembleBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 66.43 | |
| KDE (Gaussian)Backbone=ViT-B/32, Method Type=Feature-based2026.02 | 66.35 | |
| MC-DropoutBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 66.27 | |
| FeatureNormBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 66.03 | |
| ResidualBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 66.03 | |
| GradNormBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 65.94 | |
| PCXBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 65.87 | |
| GRAMBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 64.9 | |
| GradOrthBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 64.22 | |
| NANBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 64.1 | |
| PCXBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 64.1 | |
| WeiPerBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 64.04 | |
| ODINBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 63.97 | |
| KDE (Gaussian)Backbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 63.85 | |
| MCPBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 63.72 | |
| LOFBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 63.01 | |
| GradNormBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 61.94 | |
| ASHBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 61.25 | |
| NANBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 60.9 | |
| NMDBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 60.51 | |
| ASHBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 60.45 | |
| Deep EnsembleBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 60.32 | |
| GradOrthBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 59.46 | |
| TAPUUDBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 59.36 | |
| WeiPerBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 58.85 | |
| ODINBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 58.46 | |
| MC-DropoutBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 58.32 | |
| DICEBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 58.21 | |
| MCPBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 57.95 | |
| GRAMBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 57.55 | |
| DICEBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 57.05 | |
| SHEBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 56.54 | |
| NuSABackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 55.64 | |
| ViMBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 55.45 | |
| ReActBackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 54.87 | |
| ReActBackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 54.36 | |
| TAPUUDBackbone=ResNet18, Method Category=Feature-based, Hyperparameter Setting=best-performing2026.02 | 53.14 | |
| FeatureNormBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 52.28 | |
| NACBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 52.1 | |
| GAIA-ABackbone=ViT-B/32, Method Type=Confidence-based2026.02 | 52.05 | |
| NANBackbone=VGG16, Method category=Feature-based, Hyperparameter selection=best-performing2026.02 | 51.77 | |
| GAIA-ABackbone=VGG16, Method category=Confidence-based, Hyperparameter selection=best-performing2026.02 | 48.33 | |
| NACBackbone=ViT-B/32, Method Type=Feature-based2026.02 | 48.29 | |
| GAIA-ABackbone=ResNet18, Method Category=Confidence-based, Hyperparameter Setting=best-performing2026.02 | 36.47 |