{"id":"W2109886035","doi":"10.1109/icassp.2016.7472618","title":"End-to-end attention-based large vocabulary speech recognition","year":2016,"lang":"en","type":"preprint","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":124,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"Compute Canada; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Canadian Institute for Advanced Research","keywords":"Computer science; Hidden Markov model; Recurrent neural network; Speech recognition; Decoding methods; Vocabulary; Sequence (biology); Pooling; Artificial intelligence; Character (mathematics); Language model; Process (computing); Pattern recognition (psychology); Artificial neural network; Algorithm","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0009400473,0.0004565143,0.0004547439,0.0006456989,0.0001730486,0.0004482972,0.001403332,0.0004356206,0.006774079],"category_scores_gemma":[0.0002100452,0.0003816348,0.0004287359,0.0003386169,0.00003928691,0.0003162272,0.001011658,0.0004167266,0.008813525],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001585513,"about_ca_system_score_gemma":0.000319923,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004832787,"about_ca_topic_score_gemma":0.00006294052,"domain_scores_codex":[0.9964646,0.000255365,0.000575304,0.001315938,0.0007738587,0.0006149127],"domain_scores_gemma":[0.9972631,0.0003371029,0.0002362938,0.001393853,0.0004136615,0.0003559413],"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":[0.00001619125,0.0002700356,0.0003434989,0.00007965991,0.00007819768,0.0001057536,0.00003614506,0.00000118415,0.001006412,0.001274076,0.01402768,0.9827611],"study_design_scores_gemma":[0.007212972,0.0006484445,0.02527181,0.005724936,0.0004498568,0.0002870476,0.0001801727,0.07030804,0.4177328,0.2745286,0.187609,0.01004631],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008313604,0.00003578897,0.9217717,0.007686906,0.002112836,0.0007185619,0.0002774925,0.0009256444,0.05815744],"genre_scores_gemma":[0.2383082,0.00006209214,0.7377222,0.0117028,0.00105101,0.0005285342,0.0005862279,0.0001150406,0.00992391],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9727148,"threshold_uncertainty_score":0.9998636,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03520127758503985,"score_gpt":0.2663043535830252,"score_spread":0.2311030759979854,"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."}}