{"id":"https://openalex.org/W4372259859","doi":"https://doi.org/10.1109/icassp49357.2023.10094609","title":"Lattice-Free Sequence Discriminative Training for Phoneme-Based Neural Transducers","display_name":"Lattice-Free Sequence Discriminative Training for Phoneme-Based Neural Transducers","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4372259859","doi":"https://doi.org/10.1109/icassp49357.2023.10094609"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10094609","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10094609","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5055490509","display_name":"Zijian Gy\u0151z\u0151 Yang","orcid":"https://orcid.org/0000-0001-9955-860X"},"institutions":[{"id":"https://openalex.org/I887968799","display_name":"RWTH Aachen University","ror":"https://ror.org/04xfq0f34","country_code":"DE","type":"education","lineage":["https://openalex.org/I887968799"]}],"countries":["DE"],"is_corresponding":true,"raw_author_name":"Zijian Yang","raw_affiliation_strings":["RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074"],"affiliations":[{"raw_affiliation_string":"RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","institution_ids":["https://openalex.org/I887968799"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100636970","display_name":"Wei Zhou","orcid":"https://orcid.org/0009-0006-3754-8872"},"institutions":[{"id":"https://openalex.org/I887968799","display_name":"RWTH Aachen University","ror":"https://ror.org/04xfq0f34","country_code":"DE","type":"education","lineage":["https://openalex.org/I887968799"]},{"id":"https://openalex.org/I99977706","display_name":"FH Aachen","ror":"https://ror.org/04tqgg260","country_code":"DE","type":"education","lineage":["https://openalex.org/I99977706"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Wei Zhou","raw_affiliation_strings":["RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","AppTek GmbH, Aachen, Germany"],"affiliations":[{"raw_affiliation_string":"RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","institution_ids":["https://openalex.org/I887968799"]},{"raw_affiliation_string":"AppTek GmbH, Aachen, Germany","institution_ids":["https://openalex.org/I99977706"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088968292","display_name":"Ralf Schl\u00fcter","orcid":"https://orcid.org/0000-0003-2839-9247"},"institutions":[{"id":"https://openalex.org/I887968799","display_name":"RWTH Aachen University","ror":"https://ror.org/04xfq0f34","country_code":"DE","type":"education","lineage":["https://openalex.org/I887968799"]},{"id":"https://openalex.org/I99977706","display_name":"FH Aachen","ror":"https://ror.org/04tqgg260","country_code":"DE","type":"education","lineage":["https://openalex.org/I99977706"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Ralf Schl\u00fcter","raw_affiliation_strings":["RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","AppTek GmbH, Aachen, Germany"],"affiliations":[{"raw_affiliation_string":"RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","institution_ids":["https://openalex.org/I887968799"]},{"raw_affiliation_string":"AppTek GmbH, Aachen, Germany","institution_ids":["https://openalex.org/I99977706"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112501010","display_name":"Hermann Ney","orcid":null},"institutions":[{"id":"https://openalex.org/I887968799","display_name":"RWTH Aachen University","ror":"https://ror.org/04xfq0f34","country_code":"DE","type":"education","lineage":["https://openalex.org/I887968799"]},{"id":"https://openalex.org/I99977706","display_name":"FH Aachen","ror":"https://ror.org/04tqgg260","country_code":"DE","type":"education","lineage":["https://openalex.org/I99977706"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Hermann Ney","raw_affiliation_strings":["RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","AppTek GmbH, Aachen, Germany"],"affiliations":[{"raw_affiliation_string":"RWTH Aachen University,Human Language Technology and Pattern Recognition,Computer Science Department,Aachen,Germany,52074","institution_ids":["https://openalex.org/I887968799"]},{"raw_affiliation_string":"AppTek GmbH, Aachen, Germany","institution_ids":["https://openalex.org/I99977706"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5055490509"],"corresponding_institution_ids":["https://openalex.org/I887968799"],"apc_list":null,"apc_paid":null,"fwci":0.5245,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.69733168,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9991000294685364,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11309","display_name":"Music and Audio Processing","score":0.9943000078201294,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6392222046852112},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.6057001948356628},{"id":"https://openalex.org/keywords/lattice","display_name":"Lattice (music)","score":0.5667223334312439},{"id":"https://openalex.org/keywords/bayes-theorem","display_name":"Bayes' theorem","score":0.5508195161819458},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5338284969329834},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.5278893709182739},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.5089333057403564},{"id":"https://openalex.org/keywords/posterior-probability","display_name":"Posterior probability","score":0.4623682498931885},{"id":"https://openalex.org/keywords/naive-bayes-classifier","display_name":"Naive Bayes classifier","score":0.44622135162353516},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.44019052386283875},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.4351866841316223},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.40874823927879333},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.33952003717422485},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.22481608390808105},{"id":"https://openalex.org/keywords/support-vector-machine","display_name":"Support vector machine","score":0.10861331224441528},{"id":"https://openalex.org/keywords/acoustics","display_name":"Acoustics","score":0.08398783206939697},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.08196848630905151}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6392222046852112},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.6057001948356628},{"id":"https://openalex.org/C2781204021","wikidata":"https://www.wikidata.org/wiki/Q6497091","display_name":"Lattice (music)","level":2,"score":0.5667223334312439},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.5508195161819458},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5338284969329834},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.5278893709182739},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5089333057403564},{"id":"https://openalex.org/C57830394","wikidata":"https://www.wikidata.org/wiki/Q278079","display_name":"Posterior probability","level":3,"score":0.4623682498931885},{"id":"https://openalex.org/C52001869","wikidata":"https://www.wikidata.org/wiki/Q812530","display_name":"Naive Bayes classifier","level":3,"score":0.44622135162353516},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.44019052386283875},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4351866841316223},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40874823927879333},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.33952003717422485},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.22481608390808105},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.10861331224441528},{"id":"https://openalex.org/C24890656","wikidata":"https://www.wikidata.org/wiki/Q82811","display_name":"Acoustics","level":1,"score":0.08398783206939697},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.08196848630905151},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49357.2023.10094609","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10094609","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Reduced inequalities","score":0.7400000095367432,"id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320311419","display_name":"Ministry of Health","ror":null},{"id":"https://openalex.org/F4320311649","display_name":"Ministry of Education","ror":"https://ror.org/036nq5137"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":36,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1828163288","https://openalex.org/W1877570817","https://openalex.org/W2009150118","https://openalex.org/W2125234026","https://openalex.org/W2127141656","https://openalex.org/W2131342762","https://openalex.org/W2171842071","https://openalex.org/W2291513470","https://openalex.org/W2327501763","https://openalex.org/W2514741789","https://openalex.org/W2888909726","https://openalex.org/W2943845043","https://openalex.org/W2953561564","https://openalex.org/W2962765220","https://openalex.org/W2962826786","https://openalex.org/W2963414781","https://openalex.org/W2963747784","https://openalex.org/W3002595344","https://openalex.org/W3008174054","https://openalex.org/W3015686596","https://openalex.org/W3016234571","https://openalex.org/W3094979069","https://openalex.org/W3097747488","https://openalex.org/W3097777922","https://openalex.org/W3160551958","https://openalex.org/W3161375121","https://openalex.org/W3163560333","https://openalex.org/W3205201903","https://openalex.org/W4200629210","https://openalex.org/W4224518768","https://openalex.org/W4225741214","https://openalex.org/W4292347840","https://openalex.org/W6638749077","https://openalex.org/W6685392375","https://openalex.org/W6841975391"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2027972911","https://openalex.org/W2537862391","https://openalex.org/W2417174640","https://openalex.org/W1973600295","https://openalex.org/W1599512561","https://openalex.org/W2330406685"],"abstract_inverted_index":{"Recently,":[0],"RNN-Transducers":[1],"have":[2],"achieved":[3],"remarkable":[4],"results":[5,96],"on":[6],"various":[7],"automatic":[8],"speech":[9],"recognition":[10],"tasks.":[11],"However,":[12],"lattice-free":[13,36,40,44,50,78,99,127],"sequence":[14],"discriminative":[15],"training":[16,37,134],"methods,":[17],"which":[18,54,89],"obtain":[19],"superior":[20],"performance":[21],"in":[22,28,107,141],"hybrid":[23],"models,":[24],"are":[25,55],"rarely":[26],"investigated":[27],"RNN-Transducers.":[29],"In":[30],"this":[31],"work,":[32],"we":[33],"propose":[34],"three":[35],"objectives,":[38,126],"namely":[39],"maximum":[41],"mutual":[42],"information,":[43],"segment-level":[45],"minimum":[46,51,123],"Bayes":[47,52,124],"risk,":[48,53],"and":[49],"used":[56],"for":[57,84],"the":[58,63,81,120],"final":[59],"posterior":[60],"output":[61],"of":[62],"phoneme-based":[64],"neural":[65],"transducer":[66],"with":[67,137],"a":[68,113,138],"limited":[69],"context":[70],"dependency.":[71],"Compared":[72,118],"to":[73,91,103,112,119],"criteria":[74],"using":[75],"N-best":[76],"lists,":[77],"methods":[79,100,128],"eliminate":[80],"decoding":[82],"step":[83],"hypotheses":[85],"generation":[86],"during":[87],"training,":[88],"leads":[90],"more":[92],"efficient":[93],"training.":[94],"Experimental":[95],"show":[97],"that":[98],"gain":[101,129],"up":[102],"6.5%":[104],"relative":[105,133],"improvement":[106],"word":[108],"error":[109],"rate":[110],"compared":[111],"sequence-level":[114],"cross-entropy":[115],"trained":[116],"model.":[117],"N-best-list":[121],"based":[122],"risk":[125],"40%":[130],"-":[131],"70%":[132],"time":[135],"speedup":[136],"small":[139],"degradation":[140],"performance.":[142]},"counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}