{"id":"W2063689849","doi":"10.1109/iscslp.2012.6423452","title":"Investigation of deep neural networks (DNN) for large vocabulary continuous speech recognition: Why DNN surpasses GMMS in acoustic modeling","year":2012,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":98,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"","keywords":"Computer science; Speech recognition; Artificial neural network; Vocabulary; Context (archaeology); Mel-frequency cepstrum; Word error rate; Artificial intelligence; Task (project management); Reduction (mathematics); Deep neural networks; Hidden Markov model; Logarithm; Pattern recognition (psychology); Feature extraction; Mathematics","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":[0.002349414,0.0006890955,0.0004016277,0.00041247,0.0002760655,0.0009836054,0.0006970619,0.0009600254,0.001521621],"category_scores_gemma":[0.003884432,0.0003598653,0.0002350865,0.0005264896,0.0006025077,0.002988889,0.000605821,0.001817395,0.0005674268],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008754957,"about_ca_system_score_gemma":0.000570618,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004489874,"about_ca_topic_score_gemma":0.008039345,"domain_scores_codex":[0.999486,0.0001356057,0.00002677247,0.0001349516,0.0001656764,0.00005094539],"domain_scores_gemma":[0.9985282,0.0007938984,0.00007211812,0.000132908,0.0004178041,0.0000549449],"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":[0.0005484296,0.000205932,0.007372536,0.000464872,0.0001317376,0.0002301875,0.000242542,0.09609335,0.0563303,0.06097311,0.007448039,0.769959],"study_design_scores_gemma":[0.0000380538,0.0004518059,0.003235974,0.0001263532,0.0000652071,0.0003134511,0.0001506276,0.8885039,0.04646576,0.04155431,0.01904738,0.00004715784],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2581106,0.04416769,0.6469222,0.01492321,0.0006964692,0.0001304103,0.0003490347,0.001629015,0.0330714],"genre_scores_gemma":[0.7712332,0.01098337,0.209672,0.001181934,0.0002876875,0.00005889409,0.0002547448,0.0001503411,0.006177748],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004489874,"threshold_uncertainty_score":0.01242507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05066265927163722,"score_gpt":0.2545038198180309,"score_spread":0.2038411605463937,"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."}}