| Benchmark | Model | Rank | Results |
|---|---|---|---|
| multi-class-anomaly-detection-on-mvtec-ad | HLRAD | – | Segmentation AUROC: 98.6Segmentation AP: 71.3 |
| multi-class-anomaly-detection-on-real-iad-unified-multi-class | HLRAD | – | Detection AUROC: 90.1Segmentation AP: 47.7Detection F1-max: 81.1… |
| multi-class-anomaly-detection-on-visa-unified-multi-class | HLRAD | – | Detection AUROC: 99.0Detection F1-max: 96.6Segmentation F1-max: 57.5 |