sub:assertion {
<
https://doi.org/10.5281/zenodo.10891137>
dct:creator orcid:0000-0002-2554-3816 ,
orcid:0000-0002-5831-1237 ,
orcid:0000-0003-2175-7072 ,
orcid:0000-0003-3453-2729 ,
orcid:0000-0003-3690-3052 ;
dct:hasVersion "2.0.0" ;
dct:language <
https://www.omg.org/spec/LCC/Languages/LaISO639-1-LanguageCodes/en> ;
dct:subject <
http://purl.obolibrary.org/obo/OMIT_0026615> ;
a <
https://w3id.org/fair/ff/terms/Dataset> , <
https://w3id.org/fdof/ontology#FAIRDigitalObject> ;
rdfs:comment """Large-scale multi-label benchmark dataset for remote sensing image classification based on Sentinel-1 and Sentinel-2 imagery. Contains 549,488 pairs of image patches across 10 European countries with 19 land use/land cover classes. Each patch is 1.2 x 1.2 km with multi-spectral bands and pixel-level reference maps from CORINE Land Cover 2018. Enables deep learning research for Earth observation applications including multi-label classification, content-based image retrieval, and multi-modal learning.
LICENSING: Base Sentinel-1/2 imagery is provided under EU Regulations 377/2014 and 1159/2013 (free, full and open access for any lawful use including commercial). Dataset annotations are provided under Community Data License Agreement – Permissive, Version 1.0 (CDLA-Permissive-1.0). Attribution required: 'Contains modified Copernicus Sentinel data [2017-2018]' for imagery; citation of the BigEarthNet papers for the dataset.""" ;
rdfs:label "BigEarthNet v2.0: Large-Scale Sentinel Benchmark Archive" ;
<
https://w3id.org/fdof/ontology#hasMetadata>
this: ;
<
https://www.w3.org/ns/dcat#contactPoint> "k.clasen@tu-berlin.de" .
}