{"id":"W4408174455","doi":"10.14358/pers.24-00100r2","title":"A Comparative Study of Deep Learning Methods for Automated Road Network Extraction from High-Spatial-Resolution Remotely Sensed Imagery","year":2025,"lang":"en","type":"article","venue":"Photogrammetric Engineering & Remote Sensing","topic":"Automated Road and Building Extraction","field":"Engineering","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University; University of Calgary; University of Waterloo","funders":"","keywords":"Computer science; Deep learning; Remote sensing; Artificial intelligence; High resolution; Aerial imagery; Extraction (chemistry); Cartography; Computer vision; Geography","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006818472,0.000454784,0.0008311687,0.001061251,0.0002249109,0.0001014475,0.0001276733,0.0002740937,0.000003364513],"category_scores_gemma":[0.0005065356,0.0005192244,0.0001797907,0.002721498,0.0000234522,0.0001880111,0.00004445419,0.0006209944,0.000002456085],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003263473,"about_ca_system_score_gemma":0.00002190452,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005706045,"about_ca_topic_score_gemma":0.0001452451,"domain_scores_codex":[0.9976214,0.0002112881,0.0008296935,0.0004877652,0.0002415259,0.0006083241],"domain_scores_gemma":[0.9978318,0.001238545,0.0002555078,0.0003485495,0.0002334785,0.00009213545],"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":[0.00004478689,0.00003546929,0.00001011835,0.00007563033,0.000215908,0.000004651199,0.0002891544,0.5559301,0.1200052,0.000001039043,0.00003559841,0.3233523],"study_design_scores_gemma":[0.00106153,0.0001306184,0.006933555,0.0002505231,0.0002834791,0.00001018434,0.0003453788,0.9588625,0.03120588,0.0000408485,0.0004708363,0.0004046726],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3108402,0.0006218014,0.6836634,0.000003782801,0.001476806,0.0006800008,0.000002569523,0.002630939,0.00008052195],"genre_scores_gemma":[0.6707971,0.0000342073,0.3288913,0.000004069206,0.0001532961,0.000001988562,0.0000455779,0.00006241213,0.00001005194],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.4029324,"threshold_uncertainty_score":0.9997259,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01546688194457353,"score_gpt":0.3045463281253198,"score_spread":0.2890794461807463,"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."}}