ResNet50 + RPL

Reported on 4 benchmarks across 2 tasks · 1 paper

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

Methodology2 results

Domain AdaptationonImageNet-R
Top 1 Error· 2021-04-27
54.1
best: 4.54 (Model soups (ViT-G/14))
If your data distribution shifts, use self-learning arXiv:2104.12928
Domain AdaptationonImageNet-C
mean Corruption Error (mCE)· 2021-04-27
50.5
best: 22 (EfficientNet-L2+RPL)
If your data distribution shifts, use self-learning arXiv:2104.12928

Other2 results

Unsupervised Domain AdaptationonImageNet-R
Top 1 Error· 2021-04-27
54.1
best: 4.54 (Model soups (ViT-G/14))
If your data distribution shifts, use self-learning arXiv:2104.12928
Unsupervised Domain AdaptationonImageNet-C
mean Corruption Error (mCE)· 2021-04-27
50.5
best: 22 (EfficientNet-L2+RPL)
If your data distribution shifts, use self-learning arXiv:2104.12928