{"id":"W4415991972","doi":"10.48550/arxiv.2508.21631","title":"Towards Improved Speech Recognition through Optimized Synthetic Data Generation","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Synthetic data; Training set; Voice activity detection; Speech synthesis; Acoustic model; Encoding (memory); Pattern recognition (psychology)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002456449,0.001331966,0.0007851211,0.000664289,0.0003208974,0.0009361204,0.001466713,0.001123371,0.002568765],"category_scores_gemma":[0.007481427,0.0004881849,0.0005718278,0.0005235581,0.0007236759,0.001037838,0.001366165,0.001475103,0.001782777],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007506394,"about_ca_system_score_gemma":0.001004527,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005446059,"about_ca_topic_score_gemma":0.006602892,"domain_scores_codex":[0.998283,0.0007987021,0.00009858493,0.000360413,0.0003355721,0.0001236957],"domain_scores_gemma":[0.9958487,0.002051927,0.0001524383,0.0006346073,0.001189271,0.000123095],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000639484,0.0003586195,0.002320076,0.0003990325,0.0000998596,0.0003785913,0.000539928,0.5758843,0.1343429,0.006502124,0.009371433,0.2691637],"study_design_scores_gemma":[0.00003895018,0.0001318691,0.0004089115,0.00001481673,0.00001853956,0.0001045206,0.00006702304,0.9450072,0.04851828,0.002079858,0.00358622,0.0000238494],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06365796,0.0003547686,0.9256714,0.0003228079,0.0001508028,0.0001634682,0.0009396846,0.006839592,0.001899544],"genre_scores_gemma":[0.5049487,0.0002275705,0.4833286,0.0002156757,0.00007673363,0.0005984533,0.005817663,0.001123204,0.003663324],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005446059,"threshold_uncertainty_score":0.01299113,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1907698283837188,"score_gpt":0.3255550460538047,"score_spread":0.1347852176700859,"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."}}