{"id":"W2099621636","doi":"","title":"Vocal Tract Length Perturbation (VTLP) improves speech recognition","year":2013,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":335,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Speech recognition; Vocal tract; Normalization (sociology); Convolutional neural network; Spectrogram; Artificial neural network; Test set; Dynamic time warping; Word error rate; TIMIT; Image warping; Artificial intelligence; Pattern recognition (psychology); Hidden Markov model","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":[0.001178324,0.001370072,0.0007108885,0.0006693693,0.0002390832,0.0007533395,0.0007133445,0.0006400393,0.003504599],"category_scores_gemma":[0.004030154,0.0002854028,0.0006336476,0.0006386145,0.0004179307,0.001608846,0.001291716,0.001057808,0.00349553],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003566546,"about_ca_system_score_gemma":0.000358993,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001757143,"about_ca_topic_score_gemma":0.002692425,"domain_scores_codex":[0.9991198,0.0002002084,0.00006063042,0.0002937511,0.0002581192,0.00006748896],"domain_scores_gemma":[0.9984262,0.0006635634,0.0001282356,0.000511012,0.0002169565,0.00005407643],"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":[0.0005605828,0.0001678478,0.001957412,0.0002171746,0.00008743563,0.00016233,0.00009865159,0.02888742,0.1667939,0.000927612,0.004326576,0.7958131],"study_design_scores_gemma":[0.00006164322,0.001360471,0.01839742,0.00008504881,0.0001884989,0.0009634881,0.0001526485,0.637709,0.3128448,0.004387265,0.02372607,0.0001236396],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2150174,0.002991399,0.7426578,0.0005482757,0.0006902863,0.0001422532,0.001915378,0.02946858,0.006568536],"genre_scores_gemma":[0.6597602,0.0009818404,0.3210864,0.0002945862,0.0002438601,0.0001836959,0.007304374,0.00139063,0.008754312],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003504599,"threshold_uncertainty_score":0.01172405,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0241960568558066,"score_gpt":0.2242876426069468,"score_spread":0.2000915857511402,"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."}}