OOD Detection on CIFAR-10 IND CIFAR-100 OOD (test)
0.9037AUROCNGC
Evaluation Results
| Method | Links | |
|---|---|---|
| NGCSupervised detection=false, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.9037 | |
| F-EDLArchitecture=VGG-16, Uncertainty Type=epistemic, Setting=classical2025.10 | 0.8637 | |
| DAEDLArchitecture=VGG-16, Uncertainty Type=epistemic, Setting=classical2025.10 | 0.8604 | |
| R-EDLArchitecture=VGG-16, Uncertainty Type=epistemic, Setting=classical2025.10 | 0.8526 | |
| I-EDLArchitecture=VGG-16, Uncertainty Type=epistemic, Setting=classical2025.10 | 0.8215 | |
| EDLArchitecture=VGG-16, Uncertainty Type=epistemic, Setting=classical2025.10 | 0.8063 | |
| MSPSupervised detection=true, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.6991 | |
| SSDSupervised detection=false, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.6842 | |
| ODINSupervised detection=true, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.654 | |
| MDSupervised detection=true, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.6445 | |
| RotSupervised detection=false, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.6384 | |
| RotSupervised detection=true, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.6025 | |
| SSDSupervised detection=true, IND noise level=50% sym., OOD samples in training set=20k, OOD samples in test set=10k2021.08 | 0.5588 |