{"id":"W4390871464","doi":"10.1109/lgrs.2024.3354560","title":"Neighborhood Attention Makes the Encoder of ResUNet Stronger for Accurate Road Extraction","year":2024,"lang":"en","type":"article","venue":"IEEE Geoscience and Remote Sensing Letters","topic":"Automated Road and Building Extraction","field":"Engineering","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; Simon Fraser University","funders":"","keywords":"Computer science; Segmentation; Artificial intelligence; Deep learning; Encoder; Transformer; Residual; Image segmentation; Feature extraction; Computer vision; 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":[0.0002707392,0.0009016795,0.0004694521,0.0004415894,0.0003091021,0.0006765466,0.0009817516,0.0006844501,0.008003616],"category_scores_gemma":[0.00182793,0.0002853207,0.0004036947,0.0004001906,0.0004489387,0.001793472,0.0009267088,0.001217462,0.002356431],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005325383,"about_ca_system_score_gemma":0.0008358685,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009465438,"about_ca_topic_score_gemma":0.0233259,"domain_scores_codex":[0.9998102,0.00001194614,0.000008478542,0.00009159245,0.00004351448,0.00003440234],"domain_scores_gemma":[0.9997393,0.00006247449,0.00002144088,0.00008209595,0.00007445659,0.00002008602],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0003030636,0.0002018756,0.003175806,0.0002870841,0.00006313747,0.0003094533,0.0001305318,0.08755578,0.04551037,0.01948399,0.03253951,0.8104393],"study_design_scores_gemma":[0.00003196153,0.0001153498,0.001771861,0.00005005594,0.00006116161,0.0002564171,0.00004676775,0.9178243,0.04549153,0.01527166,0.01905316,0.00002578778],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1388161,0.001507309,0.806022,0.00125969,0.0007016239,0.0002010489,0.002550034,0.01675977,0.03218244],"genre_scores_gemma":[0.7608039,0.0006686048,0.2109896,0.0006766162,0.0001404263,0.0001203058,0.004096205,0.001104066,0.02140034],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009465438,"threshold_uncertainty_score":0.02677476,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0128991957290784,"score_gpt":0.2555446103205247,"score_spread":0.2426454145914463,"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."}}