RandAugment: Practical automated data augmentation with a reduced search space

Open paper
Benchmark Model Rank Results
data-augmentation-on-imagenetResNet-50 (RA)#14Accuracy (%): 77.6
domain-generalization-on-vizwiz-classificationEfficientNet-B7 (randaug)#17Accuracy - All Images: 45Accuracy - Corrupted Images: 38.9…
domain-generalization-on-vizwiz-classificationEfficientNet-B5 (randaug)#25Accuracy - All Images: 42.1Accuracy - Corrupted Images: 35.5…
image-classification-on-imagenetEfficientNet-B8 (RandAugment)#243Top 1 Accuracy: 85.4%
image-classification-on-imagenetEfficientNet-B7 (RandAugment)#281Top 1 Accuracy: 85%Number of params: 66M

Rank counts only results with a code link.