Out Of Distribution Detection On Imagenet 1K 12
Métriques
AUROC
FPR95
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | AUROC | FPR95 |
|---|---|---|
| mos-towards-scaling-out-of-distribution | 90.11 | 39.97 |
| extremely-simple-activation-shaping-for-out | 95.12 | 22.8 |
| non-parametric-outlier-synthesis | 91.22 | 37.93 |
| line-out-of-distribution-detection-by | 95.03 | 20.70 |
| nearest-neighbor-guidance-for-out-of-1 | 95.42 | 17.97 |
| detecting-out-of-distribution-data-through-in | - | 70.12 |
| out-of-distribution-detection-with-implicit | 83.54 | 59.83 |
| decoupling-maxlogit-for-out-of-distribution | - | 54.74 |
| scaling-for-training-time-and-post-hoc-out-of | 95.71 | 20.05 |
| rankfeat-rank-1-feature-removal-for-out-of | 92.15 | 36.8 |
| rethinking-out-of-distribution-ood-detection | 89.1 | - |
| unleashing-mask-explore-the-intrinsic-out-of | 89.24 | 40.94 |
| nearest-neighbor-guidance-for-out-of-1 | 95.45 | 19.72 |
| on-the-effectiveness-of-sparsification-for | 93.4 | 27.25 |
| boosting-out-of-distribution-detection-with | 94.28 | 27.11 |
| delving-into-out-of-distribution-detection | 91.49 | 38.17 |