OOD Detection on NINCO (test)
94.98AUROCX-Maha†
Evaluation results
| Method | Links | ||
|---|---|---|---|
| X-Maha†Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 94.98 | 26.74 | |
| MM++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 91.34 | 43.2 | |
| rMaha++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 90.48 | 44.33 | |
| Maha++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 90.37 | 47.3 | |
| rMahaBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 88.51 | 57.85 | |
| MahaBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 85.77 | 70.01 | |
| ReActBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 85.62 | 48.58 | |
| MSPBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 83.28 | 61.85 | |
| EnergyBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 83.23 | 51.22 | |
| ODINBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 74.28 | 64.69 | |
| KNNBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 40.44 | 96 |