Image Classification on CIFAR-100 (Top-5 Accuracy)
86.26Top-5 AccuracyOurs
Evaluation results
| Method | Links | |
|---|---|---|
| OursDistortion Type=Random missing pixels (RM), Evaluation Protocol=Linear probing, Pre-trained on=ImageNet-1002021.10 | 86.26 | |
| BaselineDistortion Type=Random missing pixels (RM), Evaluation Protocol=Linear probing, Pre-trained on=ImageNet-1002021.10 | 84.09 | |
| OursDistortion Type=Additive Gaussian noise (GN), Evaluation Protocol=Linear probing, Pre-trained on=ImageNet-1002021.10 | 81.5 | |
| BaselineDistortion Type=Additive Gaussian noise (GN), Evaluation Protocol=Linear probing, Pre-trained on=ImageNet-1002021.10 | 81.37 |