OOD Detection on Bravo-Synobj
99AUROCOurs (SwinB-NF)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Ours (SwinB-NF)Category=VFMs, Backbone=SwinB, Modeling=NF2025.01 | 99 | 2 | |
| Ours (RN50-NF)Category=VFMs, Backbone=RN50, Modeling=NF2025.01 | 92 | 57 | |
| SwinB-NFCategory=Image-Net Trained Counterparts, Backbone=SwinB, Modeling=NF2025.01 | 91 | 53 | |
| Ours (RN50-GMM)Category=VFMs, Backbone=RN50, Modeling=GMM2025.01 | 89 | 66 | |
| RN50-NFCategory=Image-Net Trained Counterparts, Backbone=ResNet-50, Modeling=NF2025.01 | 87 | 74 | |
| RN50-GMMCategory=Image-Net Trained Counterparts, Backbone=ResNet-50, Modeling=GMM2025.01 | 84 | 75 | |
| EntropyCategory=OOD Methods, Backbone=ResNet-502025.01 | 61 | 86 | |
| DICECategory=OOD Methods, Backbone=ResNet-502025.01 | 57 | 91 | |
| MaxLogitCategory=OOD Methods, Backbone=ResNet-502025.01 | 56 | 95 | |
| ODINCategory=OOD Methods, Backbone=ResNet-502025.01 | 56 | 95 | |
| EnergyCategory=OOD Methods, Backbone=ResNet-502025.01 | 50 | 97 | |
| KNNCategory=OOD Methods, Backbone=ResNet-502025.01 | 49 | 91 | |
| GradNormCategory=OOD Methods, Backbone=ResNet-502025.01 | 45 | 94 |