Boundary Data set : Land Cover Classification with Gaussian Processes using spatio-spectro-temporal features

Boundary data set used to evaluate the continuity predictions for different models in boundary zones. It is composed of labelled and unlabelled pixels for a boundary size of 100m and 200m.

For further details see section VI-A-1 of the pre-print article "Land Cover Classification with Gaussian Processes using spatio-spectro-temporal features ". This article is available here.

To compute the predictions in the boundary zones with different models (GP, RF, MLP, LTAE), the code is available in the open source repository.