Anomaly Detection on RSNA
91AU-ROC (Image-level, Det.)DDAD
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DDADBackbone=AE-U, Anomaly Score=A_inter2022.06 | 91 | — | — | — | |
| DDADBackbone=AE-U, Anomaly Score=A_intra2022.06 | 87.3 | — | — | — | |
| AE-UAnomaly Score=A_rec2022.06 | 86.7 | — | — | — | |
| InvADEvaluation Setting=single-class, FPS=882025.04 | 82.7 | 87.2 | — | — | |
| DDADBackbone=AE, Anomaly Score=A_inter2022.06 | 81.5 | — | — | — | |
| f-AnoGANAnomaly Score=A_rec2022.06 | 79.8 | — | — | — | |
| hyperbolic frameworkBackbone=WideResNet50, Feature extractor dimensionality=1024, Patch size=3, Optimizer=Adam, Learning rate=10^-3, Batch size=32, Epochs=502025.05 | 79.46 | — | — | — | |
| DDADBackbone=MemAE, Anomaly Score=A_inter2022.06 | 78.8 | — | — | — | |
| PaDiMEvaluation Setting=single-class, FPS=202025.04 | 77.5 | 79.9 | — | — | |
| PatchCoreEvaluation Setting=single-class, FPS=202025.04 | 76.1 | 86.4 | — | — | |
| PatchCoreBackbone=WideResNet502025.05 | 75.67 | — | — | — | |
| PaDiMBackbone=WideResNet502025.05 | 74.48 | — | — | — | |
| DDADBackbone=MemAE, Anomaly Score=A_intra2022.06 | 72.9 | — | — | — | |
| CFlowEvaluation Setting=single-class, FPS=152025.04 | 71.5 | 77 | — | — | |
| DDADBackbone=AE, Anomaly Score=A_intra2022.06 | 69.4 | — | — | — | |
| STFPMBackbone=WideResNet502025.05 | 69.31 | — | — | — | |
| SimpleNetEvaluation Setting=single-class, FPS=102025.04 | 69.1 | 80.7 | — | — | |
| MemAEAnomaly Score=A_rec2022.06 | 68 | — | — | — | |
| RD4ADBackbone=WideResNet502025.05 | 67.63 | — | — | — | |
| RD4ADEvaluation Setting=single-class, FPS=202025.04 | 67.6 | 84.2 | — | — | |
| AEAnomaly Score=A_rec2022.06 | 66.9 | — | — | — | |
| CFABackbone=WideResNet502025.05 | 66.65 | — | — | — | |
| FAE-SSIM2026.04 | — | — | 89.2 | — | |
| FAE-SSIM2026.04 | — | — | — | 91.1 | |
| MSDE2026.04 | — | — | 90.6 | — | |
| MSDEBackbone=AnatPaste2026.04 | — | — | — | 91.8 |