https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/Head https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://www.nanopub.org/nschema#hasAssertion https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/assertion https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://www.nanopub.org/nschema#hasProvenance https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/provenance https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://www.nanopub.org/nschema#hasPublicationInfo https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/pubinfo https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.nanopub.org/nschema#Nanopublication https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/assertion https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome http://schema.org/endDate 2026-04-17 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome http://www.w3.org/1999/02/22-rdf-syntax-ns#type https://w3id.org/sciencelive/o/terms/FORRT-Replication-Outcome https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome http://www.w3.org/2000/01/rdf-schema#label Simple Prototypical Networks transfer from photographs to Sentinel-2 at 67% accuracy https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasConclusionDescription Prototypical Networks with a simple 4-block convolutional backbone (110,000 parameters) trained on everyday photographs classify Sentinel-2 satellite land cover types at 67.4% accuracy with 5 labeled examples per class. Cross-domain transfer from photographs to satellite imagery works even with a minimal architecture, but a deeper backbone and higher image resolution are needed to reach the 73-75% accuracy range. https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasConfidenceLevel https://w3id.org/sciencelive/o/terms/HighConfidence https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasEvidenceDescription 5-way 5-shot: 67.4% ± 0.7%. 5-way 20-shot: 73.1% ± 0.6%. 5-way 50-shot: 74.7% ± 0.6%. Evaluated on EuroSAT (27,000 real Sentinel-2 satellite images, 10 land cover classes) over 600 random classification tasks. Training: 5,000 episodic steps on mini-ImageNet (38,400 photographs, 64 object categories). 4-block CNN backbone, 84×84 pixel images. https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasLimitationsDescription The 4-block CNN has 45 times fewer parameters than ResNet-10 and processes 84×84 images instead of 224×224. These architectural choices explain the lower accuracy. Only RGB bands used (3 of 13 available Sentinel-2 spectral bands). EuroSAT classes are broad land cover categories, not fine-grained habitat types. https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasOutcomeRepository https://doi.org/10.5281/zenodo.19643138 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/hasValidationStatus https://w3id.org/sciencelive/o/terms/PartiallySupported https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/o/terms/isOutcomeOf https://w3id.org/sciencelive/np/RAytukwQbTLwKSG4r7oEP7jRtgoBKTQVvO0EAzj6xEUiY/conv4-baseline https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/provenance https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/assertion http://www.w3.org/ns/prov#wasAttributedTo https://orcid.org/0000-0002-1784-2920 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/pubinfo https://orcid.org/0000-0002-1784-2920 http://xmlns.com/foaf/0.1/name Anne Fouilloux https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://purl.org/dc/terms/created 2026-04-18T16:13:06.545Z https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://purl.org/dc/terms/creator https://orcid.org/0000-0002-1784-2920 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://purl.org/dc/terms/license https://creativecommons.org/licenses/by/4.0/ https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://purl.org/nanopub/x/introduces https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/cross-domain-outcome https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://purl.org/nanopub/x/wasCreatedAt https://platform.sciencelive4all.org https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 http://www.w3.org/2000/01/rdf-schema#label NP created using Declaring a replication study outcome according to FORRT https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 https://w3id.org/np/o/ntemplate/wasCreatedFromTemplate https://w3id.org/np/RA2zljn0Nw9SadppOyxZoh-_Rxosslrq-vYG-p9SttnJE https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/sig http://purl.org/nanopub/x/hasAlgorithm RSA https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/sig http://purl.org/nanopub/x/hasPublicKey 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 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/sig http://purl.org/nanopub/x/hasSignature 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 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/sig http://purl.org/nanopub/x/hasSignatureTarget https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8 https://w3id.org/sciencelive/np/RA7OZmOmun07jDm8q6lq7Ris1W0MU3rptcp3bphOWUJj8/sig http://purl.org/nanopub/x/signedBy https://orcid.org/0000-0002-1784-2920