ResNeXt101 32x8d + DeepAug + Augmix + RPL
Reported on 2 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.
Methodology1 result
- mean Corruption Error (mCE)· 2021-04-27SOTA34.8best: 22 (EfficientNet-L2+RPL)
Other1 result
- mean Corruption Error (mCE)· 2021-04-27SOTA34.8best: 22 (EfficientNet-L2+RPL)