{"id":"W2160978380","doi":"10.1109/icassp.2007.366915","title":"Integration of Multiple Feature Sets for Reducing Ambiguity in ASR","year":2007,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"University of Edinburgh","keywords":"Mel-frequency cepstrum; Hidden Markov model; Speech recognition; Computer science; Ambiguity; Feature (linguistics); Pattern recognition (psychology); Phone; Cepstrum; Feature extraction; Artificial intelligence; Artificial neural network","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":[0.00166152,0.0008906246,0.001231712,0.0006506987,0.0002602596,0.0008239038,0.0008819527,0.0006669186,0.001192997],"category_scores_gemma":[0.004593257,0.0004632962,0.0006361515,0.000544392,0.0004327825,0.001815426,0.001002122,0.0008531379,0.0004064351],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002465881,"about_ca_system_score_gemma":0.000282274,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003579872,"about_ca_topic_score_gemma":0.0006016352,"domain_scores_codex":[0.9985219,0.0003980272,0.0001061097,0.0002068874,0.000674088,0.00009311915],"domain_scores_gemma":[0.9979265,0.001147698,0.0001887583,0.0003873947,0.0003087196,0.00004088199],"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":[0.0005655732,0.0001817757,0.0009955565,0.0001367577,0.0001043154,0.0002060469,0.0001313646,0.04710203,0.1771812,0.006369572,0.0003439323,0.766682],"study_design_scores_gemma":[0.00007935573,0.0009000218,0.003729091,0.00002779611,0.0001605828,0.0008118557,0.00006094404,0.8242326,0.1599121,0.006825038,0.003156946,0.0001037467],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07950871,0.0005028348,0.9181049,0.00009001108,0.00003456918,0.00006354706,0.00002356445,0.0006715544,0.001000407],"genre_scores_gemma":[0.5346833,0.0002123506,0.4637149,0.000058529,0.00006521821,0.00007350078,0.00008716132,0.0001182943,0.0009866654],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00166152,"threshold_uncertainty_score":0.008787096,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04468577413315181,"score_gpt":0.3061615793647963,"score_spread":0.2614758052316444,"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."}}