{"id":"W4221166146","doi":"10.1016/j.jvcir.2023.103800","title":"TransCAM: Transformer attention-based CAM refinement for Weakly supervised semantic segmentation","year":2023,"lang":"en","type":"article","venue":"Journal of Visual Communication and Image Representation","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":85,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto; University of New Brunswick","funders":"","keywords":"Computer science; Artificial intelligence; Transformer; Segmentation; Convolutional neural network; Discriminative model; Pattern recognition (psychology); Pixel; Pascal (unit); Voltage","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":[0.0008921815,0.002355631,0.002800244,0.002320876,0.0009116994,0.001793568,0.00430308,0.002640227,0.015162],"category_scores_gemma":[0.002919604,0.001101224,0.001926552,0.001886022,0.0009102689,0.002237019,0.002844827,0.002747361,0.005317423],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009610242,"about_ca_system_score_gemma":0.002009028,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01467683,"about_ca_topic_score_gemma":0.03826694,"domain_scores_codex":[0.9993297,0.00008197228,0.00003581217,0.0002447063,0.0001952387,0.000112502],"domain_scores_gemma":[0.999005,0.0003335944,0.00005816761,0.0002768571,0.0002603592,0.00006610101],"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":[0.0007501912,0.0002380739,0.0006165539,0.0003235008,0.0001715428,0.0002310827,0.0001748383,0.03778218,0.05362326,0.009014331,0.02554277,0.8715315],"study_design_scores_gemma":[0.00004954257,0.00008949159,0.0003061702,0.00003365845,0.00004952187,0.0001544194,0.00005063238,0.9617195,0.02188266,0.009905807,0.005730497,0.00002817133],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009773324,0.0005697993,0.9690954,0.0001339345,0.0001440616,0.0001674437,0.0006165298,0.01715861,0.002340926],"genre_scores_gemma":[0.1573884,0.0003953251,0.8252302,0.0004962182,0.0001355612,0.0002660331,0.00390546,0.003357686,0.008825078],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.015162,"threshold_uncertainty_score":0.05072188,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04025519472113439,"score_gpt":0.3715744347563467,"score_spread":0.3313192400352123,"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."}}