Automatic Data Augmentation via Invariance-Constrained Learning

Open paper
Benchmark Model Rank Results
image-classification-on-svhnWide-ResNet-28-10#51Percentage correct: 98.15
image-classification-on-cifar-10Wide-ResNet-28-10#87Percentage correct: 97.85
image-classification-on-cifar-10Wide-ResNet-40-2#116Percentage correct: 97.05
image-classification-on-cifar-100Wide-ResNet-28-10#100Percentage correct: 84.89
image-classification-on-cifar-100Wide-ResNet-40-2#148Percentage correct: 81.19

Rank counts only results with a code link.