GeoClassing Dataset

Dataset containing the images and labels for the GeoClassing data used in the CVPR NAS workshop Unseen-data challenge under the codename "Sadie"

The GeoClassing dataset is comprised of 70,000 BigEarthNet samples with corresponding labels that indicate the location being captured in the image.

The data is in a channels-first format with a shape of (n, 3, 64, 64) where n is the number of samples in the corresponding set (50,000 for training, 10,000 for validation, and 10,000 for testing).

There are ten classes in the dataset, with 7000 examples of each, distributed evenly between the three subsets.

The ten classes and corresponding numerical label are as follows:

Ireland: 0,
Lithuania: 1,
Austria: 2,
Finland: 3,
Serbia: 4,
Belgium: 5,
Luxembourg: 6,
Kosovo: 7,
Switzerland: 8,
Portugal: 9