{"id":"W4391421598","doi":"10.1016/j.csl.2024.101685","title":"Objective and subjective evaluation of speech enhancement methods in the UDASE task of the 7th CHiME challenge","year":2024,"lang":"en","type":"article","venue":"Computer Speech & Language","topic":"Speech and Audio Processing","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"Google (Canada)","funders":"Agence Nationale de la Recherche; University of Sheffield","keywords":"Computer science; Speech recognition; Speech enhancement; Leverage (statistics); Task (project management); Noise (video); Distortion (music); Artificial intelligence; Domain adaptation; Noise reduction","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":[0.004862581,0.001871396,0.001126155,0.0008369624,0.0005405745,0.001376978,0.0007438227,0.001306482,0.002308383],"category_scores_gemma":[0.01096254,0.0002372966,0.0005249565,0.0003064054,0.0007710581,0.0008898281,0.001931232,0.0009058883,0.002017357],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003443292,"about_ca_system_score_gemma":0.0004955782,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001404295,"about_ca_topic_score_gemma":0.003237337,"domain_scores_codex":[0.9949273,0.001974685,0.0004310638,0.0009687413,0.001416093,0.000282185],"domain_scores_gemma":[0.9924188,0.00324485,0.0004541141,0.0007499077,0.002399492,0.0007329687],"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":[0.01042879,0.002851246,0.02124893,0.006568902,0.001553901,0.001574256,0.003025586,0.04655745,0.3953953,0.0007860996,0.05272735,0.4572823],"study_design_scores_gemma":[0.001481384,0.01843425,0.2931702,0.0008220954,0.001408153,0.005245409,0.005423777,0.2490222,0.3453405,0.002092017,0.0765193,0.001040691],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9044301,0.005355124,0.06439551,0.0004375333,0.001210332,0.001134859,0.004862826,0.004216018,0.0139576],"genre_scores_gemma":[0.9033087,0.001055718,0.05610544,0.0005338486,0.000497852,0.0009130025,0.01974673,0.0009251414,0.01691356],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.004862581,"threshold_uncertainty_score":0.02571607,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02378122438405992,"score_gpt":0.3505703286582682,"score_spread":0.3267891042742083,"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."}}