Out Of Distribution Detection On Imagenet 1K 13
Metriken
AUROC
FPR@95
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
| Paper Title | |||
|---|---|---|---|
| Forte | 98.34 ± 00.09 | 5.18 ± 00.51 | Forte : Finding Outliers with Representation Typicality Estimation |
| ViT-B-384 Mahalanobis (pre-trained on IN-21k) | 95.0 | 27.5 | In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluation |
| DisCoPatch | 94.3 | 39.0 | DisCoPatch: Batch Statistics Are All You Need For OOD Detection, But Only If You Can Trust Them |
| EffNetv2-M Relative Mahalanobis | 88.9 | 54.9 | In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluation |
| EffNetb7 Relative Cosine Sim | 87.9 | 53.7 | In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluation |
0 of 5 row(s) selected.