{"id":"W3107092601","doi":"10.1109/access.2020.3041873","title":"DCBT-Net: Training Deep Convolutional Neural Networks With Extremely Noisy Labels","year":2020,"lang":"en","type":"article","venue":"IEEE Access","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Ministry of Science and ICT, South Korea; National Research Foundation of Korea; National IT Industry Promotion Agency; Kyungpook National University; National Research Foundation","keywords":"Computer science; Convolutional neural network; Artificial intelligence; Training (meteorology); Net (polyhedron); Machine learning; Pattern recognition (psychology); Speech recognition; Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.001847449,0.002157756,0.001050222,0.001141207,0.0008987838,0.0010846,0.002940815,0.001554842,0.002489037],"category_scores_gemma":[0.003859677,0.0007951175,0.0008521877,0.001186742,0.0009058953,0.002270017,0.001913588,0.002714005,0.001935199],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001842377,"about_ca_system_score_gemma":0.002249926,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01366941,"about_ca_topic_score_gemma":0.02860628,"domain_scores_codex":[0.9990408,0.0001165063,0.00006054716,0.0003531272,0.0002879379,0.0001411053],"domain_scores_gemma":[0.9985391,0.0003071917,0.0001732939,0.0004341335,0.0004725555,0.00007370619],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00068534,0.0004927335,0.007299969,0.000449139,0.0002779337,0.0002852429,0.0002069305,0.1360465,0.03556626,0.005051768,0.04521158,0.7684267],"study_design_scores_gemma":[0.00004812173,0.000191161,0.001802128,0.00006846806,0.00005501185,0.0001436552,0.00006677021,0.9442477,0.03933636,0.005062488,0.008936602,0.00004156106],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.110363,0.001405203,0.8484119,0.0006981011,0.0005168112,0.0003534307,0.002305825,0.02940718,0.006538473],"genre_scores_gemma":[0.3858059,0.0007409902,0.5807949,0.001265287,0.0001253295,0.0005049077,0.01444743,0.0009692518,0.01534595],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01366941,"threshold_uncertainty_score":0.02717966,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07748383957779421,"score_gpt":0.287544198648785,"score_spread":0.2100603590709908,"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."}}