{"id":"https://openalex.org/W4294311138","doi":"https://doi.org/10.1109/taslp.2022.3203891","title":"Bayesian Neural Network Language Modeling for Speech Recognition","display_name":"Bayesian Neural Network Language Modeling for Speech Recognition","publication_year":2022,"publication_date":"2022-01-01","ids":{"openalex":"https://openalex.org/W4294311138","doi":"https://doi.org/10.1109/taslp.2022.3203891"},"language":"en","primary_location":{"id":"doi:10.1109/taslp.2022.3203891","is_oa":false,"landing_page_url":"https://doi.org/10.1109/taslp.2022.3203891","pdf_url":null,"source":{"id":"https://openalex.org/S4210169297","display_name":"IEEE/ACM Transactions on Audio Speech and Language Processing","issn_l":"2329-9290","issn":["2329-9290","2329-9304"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE/ACM Transactions on Audio, Speech, and Language Processing","raw_type":"journal-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/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":["Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China"],"raw_orcid":"https://orcid.org/0000-0003-0731-3977","affiliations":[{"raw_affiliation_string":"Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China","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/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Shoukang Hu","raw_affiliation_strings":["Nanyang Technological University, Singapore"],"raw_orcid":"https://orcid.org/0000-0002-3345-6923","affiliations":[{"raw_affiliation_string":"Nanyang Technological University, Singapore","institution_ids":["https://openalex.org/I172675005"]}]},{"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":["Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China"],"raw_orcid":"https://orcid.org/0000-0002-9650-0290","affiliations":[{"raw_affiliation_string":"Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China","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":["Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China"],"raw_orcid":"https://orcid.org/0000-0002-7886-439X","affiliations":[{"raw_affiliation_string":"Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China","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":["Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China"],"raw_orcid":"https://orcid.org/0000-0001-6725-1160","affiliations":[{"raw_affiliation_string":"Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China","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":["Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Hong Kong, SAR, China","institution_ids":["https://openalex.org/I177725633"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.2803,"has_fulltext":false,"cited_by_count":30,"citation_normalized_percentile":{"value":0.93025123,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":89,"max":99},"biblio":{"volume":"30","issue":null,"first_page":"2900","last_page":"2917"},"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":0.9959999918937683,"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":0.9879999756813049,"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/perplexity","display_name":"Perplexity","score":0.8276214599609375},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7769296765327454},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.7588376402854919},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6614352464675903},{"id":"https://openalex.org/keywords/overfitting","display_name":"Overfitting","score":0.5483893752098083},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5180971026420593},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5170264840126038},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5080235004425049},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.5073928236961365},{"id":"https://openalex.org/keywords/bayesian-inference","display_name":"Bayesian inference","score":0.48139241337776184},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.4805898666381836},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.46857619285583496},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.40347665548324585},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.40077388286590576}],"concepts":[{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.8276214599609375},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7769296765327454},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.7588376402854919},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6614352464675903},{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.5483893752098083},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5180971026420593},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5170264840126038},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5080235004425049},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.5073928236961365},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.48139241337776184},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.4805898666381836},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.46857619285583496},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40347665548324585},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.40077388286590576},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/taslp.2022.3203891","is_oa":false,"landing_page_url":"https://doi.org/10.1109/taslp.2022.3203891","pdf_url":null,"source":{"id":"https://openalex.org/S4210169297","display_name":"IEEE/ACM Transactions on Audio Speech and Language Processing","issn_l":"2329-9290","issn":["2329-9290","2329-9304"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE/ACM Transactions on Audio, Speech, and Language Processing","raw_type":"journal-article"},{"id":"pmh:oai:dr.ntu.edu.sg:10356/164438","is_oa":false,"landing_page_url":"https://hdl.handle.net/10356/164438","pdf_url":null,"source":{"id":"https://openalex.org/S4306402609","display_name":"DR-NTU (Nanyang Technological University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I172675005","host_organization_name":"Nanyang Technological University","host_organization_lineage":["https://openalex.org/I172675005"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Journal Article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/1","score":0.4099999964237213,"display_name":"No poverty"}],"awards":[{"id":"https://openalex.org/G6461416196","display_name":"Development of a Bilingual Speech Recognition System for the Elderly and Disabled Population with Speech Disorders","funder_award_id":"ITS/254/19","funder_id":"https://openalex.org/F4320326427","funder_display_name":"Innovation and Technology Fund"},{"id":"https://openalex.org/G7779516141","display_name":null,"funder_award_id":"InP/057/21","funder_id":"https://openalex.org/F4320326427","funder_display_name":"Innovation and Technology Fund"}],"funders":[{"id":"https://openalex.org/F4320326427","display_name":"Innovation and Technology Fund","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":103,"referenced_works":["https://openalex.org/W71499226","https://openalex.org/W179875071","https://openalex.org/W1495076553","https://openalex.org/W1544131626","https://openalex.org/W1567512734","https://openalex.org/W1932968309","https://openalex.org/W1934041838","https://openalex.org/W1970689298","https://openalex.org/W1985258458","https://openalex.org/W1987238397","https://openalex.org/W2052250648","https://openalex.org/W2064675550","https://openalex.org/W2076094076","https://openalex.org/W2079623482","https://openalex.org/W2094147890","https://openalex.org/W2095705004","https://openalex.org/W2111051539","https://openalex.org/W2111406701","https://openalex.org/W2111935653","https://openalex.org/W2114016253","https://openalex.org/W2127141656","https://openalex.org/W2131342762","https://openalex.org/W2135046866","https://openalex.org/W2145249131","https://openalex.org/W2150769028","https://openalex.org/W2158195707","https://openalex.org/W2194775991","https://openalex.org/W2212676342","https://openalex.org/W2242047860","https://openalex.org/W2267186426","https://openalex.org/W2293009711","https://openalex.org/W2327501763","https://openalex.org/W2413794162","https://openalex.org/W2496955520","https://openalex.org/W2514741789","https://openalex.org/W2551572271","https://openalex.org/W2559659309","https://openalex.org/W2587848299","https://openalex.org/W2749679082","https://openalex.org/W2766678531","https://openalex.org/W2798405286","https://openalex.org/W2889191672","https://openalex.org/W2936202865","https://openalex.org/W2936481169","https://openalex.org/W2937814805","https://openalex.org/W2939173691","https://openalex.org/W2943845043","https://openalex.org/W2952613254","https://openalex.org/W2972581070","https://openalex.org/W2972942899","https://openalex.org/W2981857663","https://openalex.org/W3008191852","https://openalex.org/W3015484572","https://openalex.org/W3015766686","https://openalex.org/W3015899421","https://openalex.org/W3034309359","https://openalex.org/W3048352504","https://openalex.org/W3094852745","https://openalex.org/W3095714920","https://openalex.org/W3097966334","https://openalex.org/W3113244915","https://openalex.org/W3127424476","https://openalex.org/W3140501190","https://openalex.org/W3160348851","https://openalex.org/W3161365423","https://openalex.org/W3162587184","https://openalex.org/W3162639446","https://openalex.org/W3163368926","https://openalex.org/W3163560333","https://openalex.org/W3163793923","https://openalex.org/W3198654230","https://openalex.org/W4211049957","https://openalex.org/W4212863985","https://openalex.org/W6602928593","https://openalex.org/W6617744952","https://openalex.org/W6629687487","https://openalex.org/W6634534149","https://openalex.org/W6638749077","https://openalex.org/W6638836233","https://openalex.org/W6639216784","https://openalex.org/W6640963894","https://openalex.org/W6674330103","https://openalex.org/W6680532216","https://openalex.org/W6684488266","https://openalex.org/W6688167117","https://openalex.org/W6688648137","https://openalex.org/W6696879442","https://openalex.org/W6718683173","https://openalex.org/W6729956949","https://openalex.org/W6733596270","https://openalex.org/W6735377749","https://openalex.org/W6736999424","https://openalex.org/W6737778391","https://openalex.org/W6739901393","https://openalex.org/W6745256532","https://openalex.org/W6748587240","https://openalex.org/W6752515464","https://openalex.org/W6756887525","https://openalex.org/W6757074599","https://openalex.org/W6757204547","https://openalex.org/W6780226713","https://openalex.org/W6782045920","https://openalex.org/W6840182953"],"related_works":["https://openalex.org/W2610308580","https://openalex.org/W2169518243","https://openalex.org/W2252095989","https://openalex.org/W4322096525","https://openalex.org/W2551914602","https://openalex.org/W4281893144","https://openalex.org/W2105076537","https://openalex.org/W2787311093","https://openalex.org/W2084531783","https://openalex.org/W2902731467"],"abstract_inverted_index":{"State-of-the-art":[0],"neural":[1,13,71,108],"network":[2,100],"language":[3],"models":[4],"(NNLMs)":[5],"represented":[6],"by":[7],"long":[8],"short":[9],"term":[10],"memory":[11],"recurrent":[12],"networks":[14],"(LSTM-RNNs)":[15],"and":[16,27,60,73,83,136,150,171,185,193,205,223,227,236,246],"Transformers":[17],"are":[18,23,77,142],"becoming":[19],"highly":[20],"complex.":[21],"They":[22],"prone":[24],"to":[25,52,95,103,138,221],"overfitting":[26],"poor":[28],"generalization":[29],"when":[30],"given":[31],"limited":[32],"training":[33,135],"data.":[34],"To":[35],"this":[36,50],"end,":[37],"an":[38],"overarching":[39],"full":[40],"Bayesian":[41,105,133,244],"learning":[42],"framework":[43],"encompassing":[44],"three":[45],"methods":[46],"is":[47],"proposed":[48],"in":[49,58,132,201],"paper":[51],"account":[53],"for":[54,175],"the":[55,98,128,182,213,233],"underlying":[56],"uncertainty":[57,64],"LSTM-RNN":[59,85,184,235],"Transformer":[61,87,186,237],"LMs.":[62],"The":[63],"over":[65,181,232],"their":[66,247],"model":[67,191,241],"parameters,":[68],"choice":[69],"of":[70,114,203],"activations":[72],"hidden":[74],"output":[75],"representations":[76],"modeled":[78],"using":[79,107,159],"Bayesian,":[80],"Gaussian":[81],"Process":[82],"variational":[84],"or":[86],"LMs":[88,187,238],"respectively.":[89],"Efficient":[90],"inference":[91],"approaches":[92],"were":[93,196,230],"used":[94],"automatically":[96],"select":[97],"optimal":[99],"internal":[101],"components":[102],"be":[104,139],"learned":[106],"architecture":[109],"search.":[110],"A":[111],"minimal":[112],"number":[113],"Monte":[115],"Carlo":[116],"parameter":[117],"samples":[118],"as":[119,121],"low":[120],"one":[122],"was":[123],"also":[124],"used.":[125],"These":[126],"allow":[127],"computational":[129],"costs":[130],"incurred":[131],"NNLM":[134],"evaluation":[137],"minimized.":[140],"Experiments":[141],"conducted":[143],"on":[144,212],"two":[145],"tasks:":[146],"AMI":[147],"meeting":[148],"transcription":[149],"Oxford-BBC":[151],"LipReading":[152],"Sentences":[153],"2":[154],"(LRS2)":[155],"overlapped":[156,176],"speech":[157],"recognition":[158],"state-of-the-art":[160],"LF-MMI":[161],"trained":[162],"factored":[163],"TDNN":[164],"systems":[165],"featuring":[166],"data":[167],"augmentation,":[168],"speaker":[169],"adaptation":[170],"audio-visual":[172],"multi-channel":[173],"beamforming":[174],"speech.":[177],"Consistent":[178],"performance":[179],"improvements":[180],"baseline":[183,234],"with":[188],"point":[189],"estimated":[190],"parameters":[192],"drop-out":[194],"regularization":[195],"obtained":[197,231],"across":[198],"both":[199],"tasks":[200],"terms":[202],"perplexity":[204],"word":[206],"error":[207],"rate":[208],"(WER).":[209],"In":[210],"particular,":[211],"LRS2":[214],"data,":[215],"statistically":[216],"significant":[217],"WER":[218],"reductions":[219],"up":[220],"1.3%":[222],"1.2%":[224],"absolute":[225],"(12.1%":[226],"11.3%":[228],"relative)":[229],"respectively":[239],"after":[240],"combination":[242],"between":[243],"NNLMs":[245],"respective":[248],"baselines.":[249]},"counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":10},{"year":2024,"cited_by_count":10},{"year":2023,"cited_by_count":4},{"year":2022,"cited_by_count":1}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}
