Upload of the corrupted version of Tiny ImageNet (also known as ImageNet-200) with fixed frost-corrupted samples (see https://github.com/hendrycks/robustness/issues/60). This version can be downloaded on a server (for instance using TorchUncertainty) and is safer than the original mirror on berkeley connect (that may soon be deleted). Original work by Dan Hendrycks & Thomas Dietterich under the Apache-2.0 license. If you consider this dataset useful, please cite: @article{hendrycks2019robustness, title={Benchmarking Neural Network Robustness to Common Corruptions and Perturbations}, author={Dan Hendrycks and Thomas Dietterich}, journal={Proceedings of the International Conference on Learning Representations}, year={2019} }

I am not the author of this dataset, yet, it is released under Apache-2.0 license that allows its distribution. The Apache 2.0 license is most likely under Dan Hendrycks and Thomas Dietterich's names (but it has not been completed on GitHub). Please, do not use the citation provided by Zenodo, but the one above.