Top-1 Accuracy on ImageNet-A (OOD Recognition)
5.4Top-1 AccuracyVITO
Evaluation results
| Method | Links | |
|---|---|---|
| VITOPretraining=Video pretraining, Pretraining Dataset=VideoNet2022.10 | 5.4 | |
| DINOPretraining=Standard image pretraining, Pretraining Dataset=ImageNet2022.10 | 5 | |
| MOCLRPretraining=Standard image pretraining, Pretraining Dataset=ImageNet2022.10 | 3.7 | |
| SupervisedPretraining=Supervised, Pretraining Dataset=ImageNet2022.10 | 2.2 | |
| L2-RobustPretraining=Robust image pretraining, Pretraining Dataset=ImageNet2022.10 | 2.1 | |
| Stylized-INPretraining=Robust image pretraining, Pretraining Dataset=SIN+IN2022.10 | 2 | |
| VIVIPretraining=Video pretraining, Pretraining Dataset=YT8M2022.10 | 0.5 | |
| CycleConPretraining=Video pretraining, Pretraining Dataset=R2V22022.10 | 0.4 |