PRIME + DeepAugment (ResNet-50)
Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA
Note: results are matched by exact model name. Different papers may use the same name for different model variants.
Methodology2 results
- Top 1 Accuracy· 2021-12-27SOTA59.9best: 73.6 (FAN-L-Hybrid (IN-22k))
- mean Corruption Error (mCE)· 2021-12-2751.3best: 22 (EfficientNet-L2+RPL)
Computer Vision2 results
- Top 1 Accuracy· 2021-12-27SOTA59.9best: 73.6 (FAN-L-Hybrid (IN-22k))
- mean Corruption Error (mCE)· 2021-12-2751.3best: 28.2 (DINOv2 (ViT-g/14, frozen model, linear eval))