{"id":"W3163865502","doi":"10.1109/icassp39728.2021.9413952","title":"Multi-Dialect Speech Recognition in English Using Attention on Ensemble of Experts","year":2021,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Task (project management); Word error rate; Artificial intelligence; Speech recognition; Ensemble forecasting; Baseline (sea); Variety (cybernetics); Ensemble learning; Language model; Word (group theory); Natural language processing; Training set; Machine learning; Linguistics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002143569,0.00008822772,0.0001480733,0.0001859936,0.00003287515,0.00005329828,0.0001302667,0.00006776572,0.0002176658],"category_scores_gemma":[0.000192276,0.00008925637,0.00007729358,0.0004551311,0.00001253165,0.0003059679,0.00004804975,0.0000588392,0.00003620307],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004910092,"about_ca_system_score_gemma":0.00005205263,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009243588,"about_ca_topic_score_gemma":0.0002730803,"domain_scores_codex":[0.9990149,0.0001329881,0.00024398,0.0002560216,0.0001954235,0.0001567206],"domain_scores_gemma":[0.9992997,0.0001197182,0.00006380057,0.0002343168,0.0002406172,0.00004185214],"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":[0.00001221073,0.0005408621,0.001491243,0.00001875809,0.00001277937,0.00008720148,0.0008126532,0.000008761683,0.266138,0.0001422888,0.0001795894,0.7305557],"study_design_scores_gemma":[0.0007367634,0.00004364741,0.003708287,0.0001647683,0.000004981941,0.00003086737,0.0006274857,0.04924039,0.944679,0.0004371874,0.0001249993,0.0002016641],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9056378,0.00001960734,0.08645443,0.00006593772,0.0004295382,0.00009939988,0.000001843892,0.0000890441,0.007202428],"genre_scores_gemma":[0.6856328,0.00002132815,0.3138617,0.0002142028,0.00004570009,0.000005806226,0.000007278687,0.000006753152,0.0002044318],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.730354,"threshold_uncertainty_score":0.3639769,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.085878234083952,"score_gpt":0.2885902075299994,"score_spread":0.2027119734460474,"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."}}