Benchmarking Neural Network Robustness to Common Corruptions and Perturbations
Paper • 1903.12261 • Published
Dataset Viewer
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
CIFAR-10-C (mirror)
Unchanged copy of CIFAR-10-C by Dan Hendrycks and Thomas Dietterich, from Zenodo record 2535967, licensed under CC BY 4.0. Mirrored for the CMPE 252 project Augmentation-Test-time-Adaptation so that Colab sessions can download it quickly.
Each <corruption>.npy has shape (50000, 32, 32, 3), uint8: rows
[(s-1)*10000, s*10000) are severity s. labels.npy holds the labels.
Please cite the original paper:
D. Hendrycks and T. Dietterich. Benchmarking Neural Network Robustness to Common Corruptions and Perturbations. ICLR 2019. https://arxiv.org/abs/1903.12261
- Downloads last month
- 47