{"id":"W2133401807","doi":"10.1109/icassp.2004.1327198","title":"Content based audio classification and retrieval using joint time-frequency analysis","year":2004,"lang":"en","type":"article","venue":"","topic":"Music and Audio Processing","field":"Computer Science","cited_by":58,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Centroid; Computer science; Music information retrieval; Pattern recognition (psychology); Linear discriminant analysis; Audio signal processing; Audio signal; Artificial intelligence; Speech recognition; Time–frequency analysis; Speech coding; Computer vision","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":[0.0008008627,0.000745605,0.001082796,0.003944672,0.000505996,0.00107647,0.0009054175,0.0008901805,0.002789032],"category_scores_gemma":[0.002541619,0.0002273761,0.0007140512,0.002154133,0.0004093744,0.001778476,0.0005906451,0.0004830815,0.003633805],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002881308,"about_ca_system_score_gemma":0.0003895533,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001594617,"about_ca_topic_score_gemma":0.001751516,"domain_scores_codex":[0.9990308,0.0001063374,0.00006523372,0.0002005947,0.00048732,0.0001096708],"domain_scores_gemma":[0.9985859,0.0003980273,0.000194361,0.0002620396,0.0005018406,0.00005780043],"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":[0.0003839432,0.000197434,0.001850354,0.0001087972,0.00005206305,0.0001164936,0.00007294161,0.003515888,0.1350773,0.0008814174,0.002199646,0.8555437],"study_design_scores_gemma":[0.000197434,0.001059811,0.02631662,0.00005544605,0.0003851465,0.001987739,0.0002716219,0.6910937,0.2557565,0.006429709,0.01624793,0.0001982958],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08252212,0.001067581,0.9087695,0.0001641359,0.0001649154,0.0001522283,0.0003789382,0.004186637,0.002593901],"genre_scores_gemma":[0.3114643,0.0007078723,0.6791195,0.0001278793,0.0004107846,0.0002125891,0.001511335,0.0002552892,0.006190433],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003944672,"threshold_uncertainty_score":0.009330213,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09747529054424402,"score_gpt":0.2620883164863804,"score_spread":0.1646130259421364,"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."}}