{"id":"W4285270440","doi":"10.2139/ssrn.4146179","title":"Mouth2audio: Intelligible Audio Synthesis from Videos with Distinctive Vowel Articulation","year":2022,"lang":"en","type":"article","venue":"SSRN Electronic Journal","topic":"Speech and Audio Processing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"Simon Fraser University","funders":"Social Sciences and Humanities Research Council of Canada; Simon Fraser University","keywords":"Articulation (sociology); Vowel; Speech recognition; Linguistics; Relative articulation; Computer science; Mid vowel; Communication; Psychology; Formant; Political science; Philosophy","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0003861403,0.0008494042,0.0004755691,0.0005385076,0.0001838908,0.0006138184,0.0004648248,0.0007814011,0.01647072],"category_scores_gemma":[0.0010024,0.0001752996,0.0002678342,0.0003097866,0.0001982445,0.0004233005,0.0007867417,0.00033368,0.002636221],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001245349,"about_ca_system_score_gemma":0.0002001216,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007519553,"about_ca_topic_score_gemma":0.001743607,"domain_scores_codex":[0.9997845,0.00003313481,0.0000114874,0.00005156362,0.00009034039,0.00002903952],"domain_scores_gemma":[0.9997756,0.0001056732,0.000008699392,0.00002965924,0.00005085049,0.0000295311],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003319687,0.0001298557,0.0007402188,0.0007250804,0.00007198403,0.0007451886,0.0002063777,0.005705452,0.5460591,0.002113885,0.01024543,0.4299378],"study_design_scores_gemma":[0.0008904501,0.001805119,0.007648906,0.0001395026,0.0001276405,0.002148585,0.0003127106,0.2309396,0.6969366,0.002414351,0.05650305,0.0001334758],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2393991,0.001926866,0.7069172,0.0003786701,0.001003405,0.0005233622,0.008206381,0.02230331,0.01934167],"genre_scores_gemma":[0.5366982,0.000802598,0.4210241,0.0002312268,0.0003648131,0.0004541404,0.01240397,0.002531548,0.02548939],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01647072,"threshold_uncertainty_score":0.05509996,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005571760880496717,"score_gpt":0.1990588986180614,"score_spread":0.1934871377375647,"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."}}