{"id":"W3207200417","doi":"10.1109/igarss47720.2021.9553499","title":"Global land use / land cover with Sentinel 2 and deep learning","year":2021,"lang":"en","type":"article","venue":"","topic":"Remote-Sensing Image Classification","field":"Engineering","cited_by":1249,"is_retracted":false,"has_abstract":true,"ca_institutions":"Impact","funders":"National Geographic Society","keywords":"Geospatial analysis; Land cover; Remote sensing; Computer science; Deep learning; Satellite imagery; Big data; Cloud computing; Data science; Geomatics; Earth observation; Land use; Artificial intelligence; Satellite; Cartography; Geography; Data mining; Engineering","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003331962,0.0007850717,0.0002612974,0.0007157717,0.0002144998,0.0004901082,0.0007884544,0.000509107,0.002319545],"category_scores_gemma":[0.0007906017,0.0002388196,0.0006240467,0.0008689935,0.0003067524,0.0008080343,0.000775165,0.0007681912,0.001434361],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005545013,"about_ca_system_score_gemma":0.0006666718,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01998353,"about_ca_topic_score_gemma":0.03838693,"domain_scores_codex":[0.9997826,0.00002253486,0.000009364999,0.00008389381,0.00006197581,0.00003954946],"domain_scores_gemma":[0.9998174,0.00002397157,0.0000244247,0.00004405079,0.00006750968,0.00002267175],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003718776,0.0008028548,0.04393335,0.0003242699,0.0003727449,0.000450275,0.0002808255,0.4137268,0.0368688,0.005929811,0.1589038,0.3380347],"study_design_scores_gemma":[0.00002722877,0.00007473848,0.01100141,0.00002778862,0.00002447405,0.00006427737,0.00008646258,0.957839,0.01080104,0.003986995,0.01602914,0.00003750235],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5519429,0.001242276,0.2968688,0.001836531,0.0009255022,0.0004251958,0.08361704,0.0372859,0.02585578],"genre_scores_gemma":[0.6574888,0.0004145923,0.1984728,0.0007333849,0.0001462685,0.000260263,0.1328743,0.0009996869,0.008609834],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01998353,"threshold_uncertainty_score":0.03973442,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01002611946323105,"score_gpt":0.2014934866961274,"score_spread":0.1914673672328963,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}