Journal article Open Access

CIFAR-10-C and CIFAR-10-P

Daniel Hendrycks

This repository contains the CIFAR-10-C and CIFAR-10-P dataset from Benchmarking Neural Network Robustness to Common Corruptions and Perturbations.

In CIFAR-10-C, the first 10,000 images in each .npy are the test set images corrupted at severity 1,
and the last 10,000 images are the test set images corrupted at severity five. labels.npy is the label file for all other image files.

If you find this useful in your research, please consider citing:

    @article{hendrycks2019robustness,
      title={Benchmarking Neural Network Robustness to Common Corruptions and Perturbations},
      author={Hendrycks, Dan and Dietterich, Thomas},
      journal={Proceedings of the International Conference on Learning Representations},
      year={2019}
    }

Files (21.2 GB)
Name Size
CIFAR-10-C.tar
md5:56bf5dcef84df0e2308c6dcbcbbd8499
2.9 GB Download
CIFAR-10-P.tar
md5:125d6775afc5846ea84584d7524dedff
18.3 GB Download
3,345
5,945
views
downloads
All versions This version
Views 3,3453,345
Downloads 5,9455,945
Data volume 36.1 TB36.1 TB
Unique views 2,8422,842
Unique downloads 2,5682,568

Share

Cite as