Mixup: Beyond empirical risk minimization

The dataset used in the paper is not explicitly described, but it is mentioned that the authors used CIFAR-10, CIFAR-100, ImageNet, CUB-200-2011, and Stanford Dogs datasets.

BibTex: