{"id":"W3204877369","doi":"10.48550/arxiv.2103.09300","title":"The impact of data volume on performance of deep learning based building rooftop extraction using very high spatial resolution aerial images","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Remote Sensing and LiDAR Applications","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"China Scholarship Council; University of Waterloo","keywords":"Deep learning; Computer science; Artificial intelligence; Aerial image; Segmentation; Volume (thermodynamics); Image resolution; Spatial analysis; Aerial imagery; High resolution; Extraction (chemistry); Data mining; Machine learning; Computer vision; Pattern recognition (psychology); Remote sensing; Image (mathematics); Geography","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003120987,0.001702724,0.001016986,0.001074782,0.0005328602,0.001717098,0.001209211,0.001396859,0.001177348],"category_scores_gemma":[0.01259952,0.000609005,0.0009908727,0.001249291,0.0009816632,0.003279814,0.00155786,0.001581358,0.001069203],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007756966,"about_ca_system_score_gemma":0.0006987529,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009557205,"about_ca_topic_score_gemma":0.008994563,"domain_scores_codex":[0.9981723,0.0003552268,0.0002026084,0.0005489418,0.0005151925,0.0002056199],"domain_scores_gemma":[0.9945053,0.003474856,0.000292717,0.0008942965,0.0006882659,0.0001446078],"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":[0.001926012,0.0004962895,0.02550766,0.001117078,0.000574017,0.000582506,0.0003920428,0.5722151,0.03321835,0.001583539,0.01008032,0.3523071],"study_design_scores_gemma":[0.00007071973,0.0004220452,0.01310819,0.0001656948,0.0001080641,0.0003480682,0.0003000117,0.9265753,0.05059046,0.003484885,0.004754587,0.00007184956],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.881871,0.005482791,0.09177371,0.001228051,0.0006354892,0.000162035,0.004261274,0.008280358,0.006305185],"genre_scores_gemma":[0.9337872,0.001318865,0.04768046,0.0003239964,0.00008425689,0.00009436686,0.01405902,0.0007529446,0.001898948],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009557205,"threshold_uncertainty_score":0.01900315,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05071844918125779,"score_gpt":0.2283587619531886,"score_spread":0.1776403127719308,"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."}}