ResNet-50 (SGD, Cosine)

Reported on 4 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Adversarial4 results

Adversarial RobustnessonImageNet
Accuracy· 2021-11-10
77.4
SOTA
Are Transformers More Robust Than CNNs?arXiv:2111.05464
Adversarial RobustnessonImageNet-C
mean Corruption Error (mCE)· 2021-11-10
56.9
best: 48 (DeiT-S (AdamW, Cosine))
SOTA
Are Transformers More Robust Than CNNs?arXiv:2111.05464
Adversarial RobustnessonStylized ImageNet
Accuracy· 2021-11-10
8.4
best: 13 (DeiT-S (AdamW, Cosine))
Are Transformers More Robust Than CNNs?arXiv:2111.05464
Adversarial RobustnessonImageNet-A
Accuracy· 2021-11-10
3.3
best: 12.2 (DeiT-S (AdamW, Cosine))
Are Transformers More Robust Than CNNs?arXiv:2111.05464