Metric: AUROC (higher is better)
| # | Model↕ | AUROC▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | DHM | 100 | No | - | - | - |
| 2 | Wide ResNet 40x2 | 99.9 | Yes | An Effective Baseline for Robustness to Distribu... | 2021-05-15 | Code |
| 3 | ResNet 34 + OECC+GM | 99.7 | Yes | Outlier Exposure with Confidence Control for Out... | 2019-06-08 | Code |
| 4 | ResNet 34 + FSSD | 99.5 | No | Feature Space Singularity for Out-of-Distributio... | 2020-11-30 | Code |
| 5 | Wide ResNet 40x2 | 99.43 | No | RODD: A Self-Supervised Approach for Robust Out-... | 2022-04-06 | Code |
| 6 | ZODE-KNN | 99.12 | No | Boosting Out-of-Distribution Detection with Mult... | 2022-12-24 | Code |
| 7 | ResNet18 + APR-P | 98.1 | No | Amplitude-Phase Recombination: Rethinking Robust... | 2021-08-19 | Code |
| 8 | WRN 40-2 + OE | 97.8 | Yes | Deep Anomaly Detection with Outlier Exposure | 2018-12-11 | Code |
| 9 | WRN 40-2 (MSP Baseline) | 97.8 | No | Deep Anomaly Detection with Outlier Exposure | 2018-12-11 | Code |
| 10 | WRN 40-2 + Rotation Prediction | 96.2 | No | Using Self-Supervised Learning Can Improve Model... | 2019-06-28 | Code |