Crop type classification for each of the field boundaries from the EU field boundaries dataset. The classification was made with Sentinel-2 satellite imagery timeseries, using data from 2022-01-01 until 2022-12-31. A Transformer-based deep learning model was used. The CSV has two columns: field_boundary_id (corresponding to the field boundary ID from the EU field boundaries dataset) and classification (the predicted class). The predicted class can be one of the following 28 classes: crop type alfalfa_lucerne green_silo_maize winter_rapeseed_rape grain_maize_corn_popcorn olive_plantations winter_common_soft_wheat winter_rye flowers_ornamental_plants spring_barley pasture_meadow_grassland_grass fallow_land_not_crop soy_soybeans other_tree_wood_forest vineyards_wine_vine_rebland_grapes clover oats fresh_vegetables summer_barley winter_barley rapeseed_rape potatoes sugar_beet legumes_harvested_green spring_common_soft_wheat tree_wood_forest nuts greenhouse_foil_film orchards_fruits