OOD Detection on In-house dataset
64ID PrecisionARPL
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| ARPL2022.06 | 64 | 63 | 63.5 | 68.55 | 80.61 | 99.42 | |
| OLTR2022.06 | 63 | 62 | 62.5 | 67.42 | 70.72 | 98 | |
| Mixup2022.06 | 63 | 62 | 62.5 | 66.35 | 66.7 | 97.1 | |
| Prototype2022.06 | 63 | 62 | 62.5 | 68.82 | 74.54 | 98.04 | |
| Baseline+LS+RandAug+LRSLabel Smoothing=true, RandAugment=true, Learning Rate Scheduler=true2022.06 | 62 | 63 | 62.5 | 66.19 | 63.13 | 96.34 | |
| M-T Mixup + PrototypeStrategy=MX52022.06 | 62 | 61.5 | 61.5 | 71.1 | 82.71 | 99.59 | |
| ODIN2022.06 | 61 | 59 | 60 | 64.92 | 62.79 | 96.48 | |
| M-T MixupStrategy=MX52022.06 | 61 | 60 | 60.5 | 68.78 | 70.81 | 99.29 | |
| MC-Dropout2022.06 | 59 | 58 | 58.5 | 66.07 | 68.83 | 97.57 | |
| Baseline2022.06 | 58 | 59 | 58.5 | 65.67 | 52.9 | 73.24 |