ConvNeXt-XL (Im21k) (augmentation overlap with ImageNet-C)

Reported on 2 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.

Methodology1 result

Domain AdaptationonImageNet-C
mean Corruption Error (mCE)· uses extra data· 2022-01-10
38.8
best: 22 (EfficientNet-L2+RPL)
A ConvNet for the 2020s arXiv:2201.03545

Computer Vision1 result

Domain GeneralizationonImageNet-C
mean Corruption Error (mCE)· uses extra data· 2022-01-10
38.8
best: 28.2 (DINOv2 (ViT-g/14, frozen model, linear eval))
A ConvNet for the 2020s arXiv:2201.03545