{"id":"W2136836257","doi":"10.1109/89.979381","title":"A robust compensation strategy for extraneous acoustic variations in spontaneous speech recognition","year":2002,"lang":"en","type":"article","venue":"IEEE Transactions on Speech and Audio Processing","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Hidden Markov model; Speech recognition; Computer science; Word error rate; Pattern recognition (psychology); Pronunciation; Bayesian probability; Artificial intelligence","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002652239,0.0002410941,0.0002608106,0.0004234698,0.0004367849,0.0003694833,0.0002033948,0.0001539007,0.0002043012],"category_scores_gemma":[0.00002777415,0.0002541964,0.00008603607,0.0005614799,0.00004934226,0.0006878097,0.000001707156,0.0002550712,0.00006057947],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001018918,"about_ca_system_score_gemma":0.00005533267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002530018,"about_ca_topic_score_gemma":0.0003029241,"domain_scores_codex":[0.9983206,0.00007131165,0.0004130329,0.0005532509,0.0002637974,0.0003779791],"domain_scores_gemma":[0.9990765,0.0002967587,0.0001260144,0.0002064503,0.0001673628,0.0001268863],"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":[0.00003222771,0.000283962,0.000001546872,0.0000679244,0.00001234205,0.0001589993,0.0002983462,0.002883144,0.002368501,0.00001464595,0.00003561824,0.9938427],"study_design_scores_gemma":[0.001708645,0.0003676104,0.00009357221,0.0003333349,0.00008106489,0.003579806,0.0002295811,0.9515975,0.03781336,0.003334412,0.0001828187,0.0006782689],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02244583,0.00009834931,0.974603,0.0006095877,0.0002184937,0.0005186328,0.00003521512,0.0002292631,0.001241627],"genre_scores_gemma":[0.8611007,0.0001311175,0.1379099,0.0002756429,0.00005519664,0.00009634357,0.000006385675,0.0000224318,0.0004023306],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9931645,"threshold_uncertainty_score":0.999991,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07417596438202625,"score_gpt":0.2557498206924044,"score_spread":0.1815738563103781,"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."}}