[data request] Cifar 100 corrupted data set

• Name of dataset: Cifar 100 corrupted
• URL of dataset: https://github.com/hendrycks/robustness
• License of dataset: Apache License 2.0
• Short description of dataset and use case(s): The dataset contains corrupted versions of the Cifar 100 test set and is based on the paper Benchmarking Neural Network Robustness to Common Corruptions and Perturbations Dan Hendrycks and Thomas Dietterich. 19 different corruptions are applied to the 10000 images in Cifar 100 test set and these corruptions are available in 5 different intensities. This is a benchmark dataset used for evaluating robustness of image classifiers.

Folks who would also like to see this dataset in tensorflow/datasets, please thumbs-up so the developers can know which requests to prioritize.

And if you'd like to contribute the dataset (thank you!), see our guide to adding a dataset.

Start with docs/add_dataset.md and review the Cifar 100 corrupted dataset in the linked robustness repository. Follow the dataset contribution guide to determine the required entry points and validation steps; done means the dataset is added to tensorflow/datasets and its tests pass.