Anomaly Detection On Anomaly Detection On
评估指标
Network
ROC-AUC
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Network | ROC-AUC | Paper Title | Repository |
|---|---|---|---|---|
| SCAN Features | ResNet-18 | 90.2 | Out-of-Distribution Detection Without Class Labels | - |
| GOAD | ResNet-18 | 78.8 | Classification-Based Anomaly Detection for General Data | |
| PsudoLabels ResNet-152 | ResNet-152 | 95.7 | Out-of-Distribution Detection Without Class Labels | - |
| MeanShifted | ResNet-152 | 92.6 | Mean-Shifted Contrastive Loss for Anomaly Detection | |
| PsudoLabels ViT | ViT | 99.1 | Out-of-Distribution Detection Without Class Labels | - |
| CSI | ResNet-18 | 90.3 | CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances | |
| SSD | ResNet-34 | 96.5 | SSD: A Unified Framework for Self-Supervised Outlier Detection | |
| PsudoLabels ResNet-18 | ResNet-18 | 94.3 | Out-of-Distribution Detection Without Class Labels | - |
0 of 8 row(s) selected.