{"id":"W3196806685","doi":"10.1109/tits.2021.3101053","title":"MRSDI-CNN: Multi-Model Rail Surface Defect Inspection System Based on Convolutional Neural Networks","year":2021,"lang":"en","type":"article","venue":"IEEE Transactions on Intelligent Transportation Systems","topic":"Infrastructure Maintenance and Monitoring","field":"Engineering","cited_by":86,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor; Lakehead University","funders":"Changsha Science and Technology Project; National Key Research and Development Program of China; National Natural Science Foundation of China","keywords":"Convolutional neural network; Computer science; Train; Deep learning; Artificial intelligence; Reduction (mathematics); Detector; Range (aeronautics); Artificial neural network; Pattern recognition (psychology); Machine learning; Engineering; Mathematics; Aerospace engineering; Telecommunications","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":[0.0003145504,0.0008920783,0.0004580875,0.000795338,0.0002218405,0.0003823992,0.001487933,0.0006285243,0.001922671],"category_scores_gemma":[0.0005404163,0.0004137481,0.0004986518,0.0004033777,0.0001768237,0.0009077949,0.0008260038,0.0006754243,0.0006288448],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001111829,"about_ca_system_score_gemma":0.0008720618,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01667056,"about_ca_topic_score_gemma":0.0235697,"domain_scores_codex":[0.9998026,0.00001034447,0.000008622581,0.00007156184,0.00006627983,0.00004058039],"domain_scores_gemma":[0.9998,0.00002455504,0.00002875422,0.00003480662,0.00009316786,0.00001868088],"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":[0.0006010937,0.0004616027,0.01698277,0.0002741408,0.0002900766,0.0004706401,0.00009098317,0.1861444,0.08000808,0.002438292,0.02313637,0.6891016],"study_design_scores_gemma":[0.00001099105,0.00007752458,0.00315418,0.000009188496,0.00003219265,0.00009957761,0.000008842012,0.9795542,0.01508061,0.000441116,0.001513275,0.00001831702],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2020096,0.001263471,0.7605706,0.0004865876,0.0003203004,0.0002802997,0.002610059,0.02413862,0.008320425],"genre_scores_gemma":[0.820668,0.0004867787,0.1636056,0.0003711345,0.00004239575,0.0001709993,0.003951223,0.0001520086,0.01055193],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01667056,"threshold_uncertainty_score":0.03314704,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01775501885707867,"score_gpt":0.2238167729331894,"score_spread":0.2060617540761107,"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."}}