{"id":"W217970951","doi":"","title":"Roles of Pre-Training and Fine-Tuning in Context-Dependent DBN-HMMs for Real-World Speech Recognition","year":2010,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":193,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Hidden Markov model; Computer science; Artificial intelligence; Speech recognition; Context (archaeology); Deep belief network; Vocabulary; Deep learning; Pattern recognition (psychology)","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":[0.002368216,0.0009037516,0.0006792231,0.0002482289,0.0004271232,0.0008939229,0.001014967,0.0009694697,0.00114991],"category_scores_gemma":[0.01076601,0.0005981241,0.0002994373,0.0002687143,0.0007572976,0.001834701,0.0008233417,0.002285954,0.0004435473],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005206584,"about_ca_system_score_gemma":0.0007364769,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002723801,"about_ca_topic_score_gemma":0.005689869,"domain_scores_codex":[0.998961,0.0004882344,0.00007848669,0.0002401548,0.000145004,0.00008711986],"domain_scores_gemma":[0.9953188,0.003090937,0.0001984696,0.0008934874,0.0003424922,0.0001556473],"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":[0.001019309,0.000948187,0.00597988,0.0002653667,0.0001215214,0.0001828043,0.0003519286,0.3391735,0.1617721,0.004298137,0.001051577,0.4848358],"study_design_scores_gemma":[0.0000393183,0.0002396897,0.003631412,0.00003471152,0.00003743851,0.000111494,0.00005617072,0.900795,0.09020231,0.003559868,0.001249311,0.00004330766],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2326293,0.001007637,0.7605448,0.000394664,0.0001078977,0.0001296132,0.00007674225,0.002095452,0.003013918],"genre_scores_gemma":[0.8010703,0.0002353623,0.1970919,0.0001770892,0.00002824368,0.0001128555,0.0001499794,0.0001505796,0.0009836755],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002723801,"threshold_uncertainty_score":0.01252443,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0420777987342523,"score_gpt":0.2835390570334626,"score_spread":0.2414612582992103,"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."}}