{"id":"https://openalex.org/W3162587184","doi":"https://doi.org/10.1109/icassp39728.2021.9414046","title":"Bayesian Transformer Language Models for Speech Recognition","display_name":"Bayesian Transformer Language Models for Speech Recognition","publication_year":2021,"publication_date":"2021-05-13","ids":{"openalex":"https://openalex.org/W3162587184","doi":"https://doi.org/10.1109/icassp39728.2021.9414046","mag":"3162587184"},"language":"en","primary_location":{"id":"doi:10.1109/icassp39728.2021.9414046","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414046","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5033502057","display_name":"Boyang Xue","orcid":null},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Boyang Xue","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004643540","display_name":"Jianwei Yu","orcid":"https://orcid.org/0000-0002-2449-1436"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Jianwei Yu","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011790590","display_name":"Junhao Xu","orcid":"https://orcid.org/0000-0002-9650-0290"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Junhao Xu","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066070154","display_name":"Shansong Liu","orcid":"https://orcid.org/0000-0001-6202-5615"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Shansong Liu","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045355404","display_name":"Shoukang Hu","orcid":"https://orcid.org/0000-0002-3345-6923"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Shoukang Hu","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100341722","display_name":"Zi Ye","orcid":"https://orcid.org/0000-0003-1002-0315"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zi Ye","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003895235","display_name":"Mengzhe Geng","orcid":"https://orcid.org/0000-0002-7886-439X"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Mengzhe Geng","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037109470","display_name":"Xunying Liu","orcid":"https://orcid.org/0000-0001-6725-1160"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Xunying Liu","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5019458385","display_name":"Helen Meng","orcid":"https://orcid.org/0000-0002-4427-3532"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Helen Meng","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I177725633"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":21,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"7378","last_page":"7382"},"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/T10028","display_name":"Topic Modeling","score":0.9975000023841858,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.978600025177002,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/perplexity","display_name":"Perplexity","score":0.759095311164856},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.7177008390426636},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7171329855918884},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.6424002647399902},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5183071494102478},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.5149291753768921},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.48873305320739746},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4622340798377991},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.4225367605686188},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.336880087852478},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.11697089672088623}],"concepts":[{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.759095311164856},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.7177008390426636},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7171329855918884},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.6424002647399902},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5183071494102478},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.5149291753768921},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.48873305320739746},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4622340798377991},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.4225367605686188},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.336880087852478},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.11697089672088623},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp39728.2021.9414046","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414046","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"No poverty","id":"https://metadata.un.org/sdg/1","score":0.6200000047683716}],"awards":[],"funders":[{"id":"https://openalex.org/F4320335138","display_name":"Shun Hing Institute of Advanced Engineering","ror":"https://ror.org/00t33hh48"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":50,"referenced_works":["https://openalex.org/W71499226","https://openalex.org/W179875071","https://openalex.org/W582134693","https://openalex.org/W1524333225","https://openalex.org/W1934041838","https://openalex.org/W1959608418","https://openalex.org/W2094147890","https://openalex.org/W2095705004","https://openalex.org/W2108677974","https://openalex.org/W2171928131","https://openalex.org/W2194775991","https://openalex.org/W2242047860","https://openalex.org/W2267186426","https://openalex.org/W2413794162","https://openalex.org/W2462831000","https://openalex.org/W2514741789","https://openalex.org/W2597655663","https://openalex.org/W2613904329","https://openalex.org/W2888867175","https://openalex.org/W2904530326","https://openalex.org/W2936202865","https://openalex.org/W2943845043","https://openalex.org/W2951595529","https://openalex.org/W2963386218","https://openalex.org/W2963403868","https://openalex.org/W2964059111","https://openalex.org/W2964265128","https://openalex.org/W2970971315","https://openalex.org/W2972942899","https://openalex.org/W3015484572","https://openalex.org/W3015899421","https://openalex.org/W3024195404","https://openalex.org/W3161873458","https://openalex.org/W4385245566","https://openalex.org/W6602928593","https://openalex.org/W6607333740","https://openalex.org/W6617145748","https://openalex.org/W6631362777","https://openalex.org/W6638836233","https://openalex.org/W6640963894","https://openalex.org/W6674330103","https://openalex.org/W6676315081","https://openalex.org/W6718683173","https://openalex.org/W6735377749","https://openalex.org/W6737778391","https://openalex.org/W6739901393","https://openalex.org/W6757074599","https://openalex.org/W6776169701","https://openalex.org/W6780226713","https://openalex.org/W6795108984"],"related_works":["https://openalex.org/W2169518243","https://openalex.org/W2252095989","https://openalex.org/W4322096525","https://openalex.org/W2551914602","https://openalex.org/W4287323699","https://openalex.org/W4281893144","https://openalex.org/W2105076537","https://openalex.org/W2787311093","https://openalex.org/W3013624417","https://openalex.org/W4287826556"],"abstract_inverted_index":{"State-of-the-art":[0],"neural":[1],"language":[2],"models":[3],"(LMs)":[4],"represented":[5],"by":[6],"Transformers":[7],"are":[8,59],"highly":[9],"complex.":[10],"Their":[11],"use":[12],"of":[13,72],"fixed,":[14],"deterministic":[15],"parameter":[16,65],"estimates":[17],"fail":[18],"to":[19,26,38,61,93,146],"account":[20],"for":[21,50],"model":[22,75],"uncertainty":[23],"and":[24,28,82,98,143],"lead":[25],"over-fitting":[27],"poor":[29],"generalization":[30],"when":[31],"given":[32],"limited":[33],"training":[34],"data.":[35],"In":[36],"order":[37],"address":[39],"these":[40],"issues,":[41],"this":[42],"paper":[43],"proposes":[44],"a":[45,128,136,147],"full":[46],"Bayesian":[47],"learning":[48],"framework":[49],"Transformer":[51,74,107,137],"LM":[52,131,138],"estimation.":[53],"Efficient":[54],"variational":[55],"inference":[56],"based":[57],"approaches":[58],"used":[60],"estimate":[62],"the":[63,73,105,141],"latent":[64],"posterior":[66],"distributions":[67],"associated":[68],"with":[69,118],"different":[70],"parts":[71],"architecture":[76],"including":[77],"multi-head":[78],"self-attention,":[79],"feed":[80],"forward":[81],"embedding":[83],"layers.":[84],"Statistically":[85],"significant":[86],"word":[87],"error":[88],"rate":[89],"(WER)":[90],"reductions":[91],"up":[92],"0.5%":[94],"absolute":[95],"(3.18%":[96],"relative)":[97],"consistent":[99],"perplexity":[100],"gains":[101],"were":[102,124],"obtained":[103,126],"over":[104],"baseline":[106],"LMs":[108],"on":[109,127,140],"state-of-the-art":[110],"Switchboard":[111,142],"corpus":[112],"trained":[113,139],"LF-MMI":[114],"factored":[115],"TDNN":[116],"systems":[117],"i-Vector":[119],"speaker":[120],"adaptation.":[121],"Performance":[122],"improvements":[123],"also":[125],"cross":[129],"domain":[130],"adaptation":[132],"task":[133],"requiring":[134],"porting":[135],"Fisher":[144],"data":[145],"low-resource":[148],"DementiaBank":[149],"elderly":[150],"speech":[151],"corpus.":[152]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":6},{"year":2023,"cited_by_count":3},{"year":2022,"cited_by_count":6},{"year":2021,"cited_by_count":4}],"updated_date":"2026-07-29T09:40:50.615796","created_date":"2025-10-10T00:00:00"}