-
CIFAR-10-C and CIFAR-100-C
CIFAR-10-C and CIFAR-100-C are robustness benchmarks consisting of 19 corruptions types with five levels of severities. -
CIFAR-10-C and ImageNet-C
The dataset used in the paper is CIFAR-10-C and ImageNet-C, which are common corruption benchmarks. -
Wilds: A Benchmark of In-the-Wild Distribution Shifts
The dataset used in the paper is a collection of images for domain generalization tasks, including CIFAR-10-C, CIFAR-100-C, and Digit-DG. -
CIFAR-10-C
CIFAR-10-C is a dataset of 60,000 32x32 color images in 10 classes, with 6,000 images per class, and 10% of the images are corrupted.
