{"id":"https://openalex.org/W4280635176","doi":"https://doi.org/10.21437/interspeech.2022-11169","title":"Improved Consistency Training for Semi-Supervised Sequence-to-Sequence ASR via Speech Chain Reconstruction and Self-Transcribing","display_name":"Improved Consistency Training for Semi-Supervised Sequence-to-Sequence ASR via Speech Chain Reconstruction and Self-Transcribing","publication_year":2022,"publication_date":"2022-09-16","ids":{"openalex":"https://openalex.org/W4280635176","doi":"https://doi.org/10.21437/interspeech.2022-11169"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2022-11169","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-11169","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2022","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/A5038740801","display_name":"Heli Qi","orcid":"https://orcid.org/0000-0001-9512-7140"},"institutions":[{"id":"https://openalex.org/I75917431","display_name":"Nara Institute of Science and Technology","ror":"https://ror.org/05bhada84","country_code":"JP","type":"education","lineage":["https://openalex.org/I75917431"]}],"countries":["JP"],"is_corresponding":true,"raw_author_name":"Heli Qi","raw_affiliation_strings":["Nara Institute of Science and Technology, Japan"],"affiliations":[{"raw_affiliation_string":"Nara Institute of Science and Technology, Japan","institution_ids":["https://openalex.org/I75917431"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048192694","display_name":"Sashi Novitasari","orcid":"https://orcid.org/0000-0001-7467-5682"},"institutions":[{"id":"https://openalex.org/I75917431","display_name":"Nara Institute of Science and Technology","ror":"https://ror.org/05bhada84","country_code":"JP","type":"education","lineage":["https://openalex.org/I75917431"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Sashi Novitasari","raw_affiliation_strings":["Nara Institute of Science and Technology, Japan"],"affiliations":[{"raw_affiliation_string":"Nara Institute of Science and Technology, Japan","institution_ids":["https://openalex.org/I75917431"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040108974","display_name":"Sakriani Sakti","orcid":"https://orcid.org/0000-0001-5509-8963"},"institutions":[{"id":"https://openalex.org/I177738480","display_name":"Japan Advanced Institute of Science and Technology","ror":"https://ror.org/03frj4r98","country_code":"JP","type":"education","lineage":["https://openalex.org/I177738480"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Sakriani Sakti","raw_affiliation_strings":["Japan Advanced Institute of Science and Technology, Japan"],"affiliations":[{"raw_affiliation_string":"Japan Advanced Institute of Science and Technology, Japan","institution_ids":["https://openalex.org/I177738480"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5020994673","display_name":"Satoshi Nakamura","orcid":"https://orcid.org/0000-0001-6956-3803"},"institutions":[{"id":"https://openalex.org/I75917431","display_name":"Nara Institute of Science and Technology","ror":"https://ror.org/05bhada84","country_code":"JP","type":"education","lineage":["https://openalex.org/I75917431"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Satoshi Nakamura","raw_affiliation_strings":["Nara Institute of Science and Technology, Japan"],"affiliations":[{"raw_affiliation_string":"Nara Institute of Science and Technology, Japan","institution_ids":["https://openalex.org/I75917431"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5038740801"],"corresponding_institution_ids":["https://openalex.org/I75917431"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.03275156,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"3413","last_page":"3417"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9995999932289124,"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":0.9995999932289124,"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":0.9872000217437744,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9757999777793884,"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/sequence","display_name":"Sequence (biology)","score":0.7385127544403076},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7168562412261963},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.6146734356880188},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5558736324310303},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.5424980521202087},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5385554432868958},{"id":"https://openalex.org/keywords/biology","display_name":"Biology","score":0.05156201124191284}],"concepts":[{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.7385127544403076},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7168562412261963},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.6146734356880188},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5558736324310303},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5424980521202087},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5385554432868958},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.05156201124191284},{"id":"https://openalex.org/C54355233","wikidata":"https://www.wikidata.org/wiki/Q7162","display_name":"Genetics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2022-11169","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-11169","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2022","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.6600000262260437,"display_name":"Quality Education"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320334764","display_name":"Japan Society for the Promotion of Science","ror":"https://ror.org/00hhkn466"},{"id":"https://openalex.org/F4320338335","display_name":"H2020 European Research Council","ror":"https://ror.org/0472cxd90"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W6908809","https://openalex.org/W854541894","https://openalex.org/W1494198834","https://openalex.org/W1522301498","https://openalex.org/W2129068307","https://openalex.org/W2133564696","https://openalex.org/W2327501763","https://openalex.org/W2431080869","https://openalex.org/W2890964092","https://openalex.org/W2936774411","https://openalex.org/W2951970475","https://openalex.org/W2962699523","https://openalex.org/W2962780374","https://openalex.org/W2962826786","https://openalex.org/W2964243274","https://openalex.org/W2978426779","https://openalex.org/W3001197829","https://openalex.org/W3015280134","https://openalex.org/W3015737168","https://openalex.org/W3026041220","https://openalex.org/W3096273170","https://openalex.org/W3097647528","https://openalex.org/W3161101519","https://openalex.org/W3167533889","https://openalex.org/W3198840231","https://openalex.org/W4287118210"],"related_works":["https://openalex.org/W1603736412","https://openalex.org/W4304185162","https://openalex.org/W2061685118","https://openalex.org/W3006282800","https://openalex.org/W2462100143","https://openalex.org/W1770503332","https://openalex.org/W3125207769","https://openalex.org/W2577825108","https://openalex.org/W2949684406","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Consistency":[0],"regularization":[1],"has":[2],"recently":[3],"been":[4],"applied":[5],"to":[6,19,48,58,94,137,166],"semi-supervised":[7,34,126],"sequence-to-sequence":[8],"(S2S)":[9],"automatic":[10],"speech":[11,27,90,130],"recognition":[12],"(ASR).This":[13],"principle":[14],"encourages":[15],"an":[16,120],"ASR":[17,36,86,152],"model":[18,47,107,153],"output":[20],"similar":[21],"predictions":[22],"for":[23,52],"the":[24,65,72,76,80,96,104,134,150,155,178,184],"same":[25],"input":[26,89],"with":[28],"different":[29],"perturbations.The":[30],"existing":[31],"paradigm":[32,56,124,171],"of":[33,62,68,75,82,125],"S2S":[35,85,127],"utilizes":[37],"SpecAugment":[38,69],"as":[39,133],"data":[40],"augmentation":[41,136],"and":[42,91,160,181],"requires":[43,87],"a":[44,173],"static":[45,99],"teacher":[46,106],"produce":[49],"pseudo":[50,83,111,140,146],"transcripts":[51,147],"untranscribed":[53],"speech.However,":[54],"this":[55,116],"fails":[57],"take":[59],"full":[60],"advantage":[61],"consistency":[63,114,122,156],"regularization.First,":[64],"masking":[66],"operations":[67],"may":[70],"damage":[71],"linguistic":[73],"contents":[74],"speech,":[77],"thus":[78],"influencing":[79],"quality":[81],"labels.Second,":[84],"both":[88],"prefix":[92,100],"tokens":[93,101],"make":[95],"next":[97],"prediction.The":[98],"made":[102],"by":[103,149],"offline":[105],"cannot":[108],"match":[109],"dynamic":[110,145],"labels":[112],"during":[113],"training.In":[115],"work,":[117],"we":[118,142],"propose":[119],"improved":[121,170],"training":[123],"ASR.We":[128],"utilize":[129],"chain":[131],"reconstruction":[132],"weak":[135],"generate":[138],"high-quality":[139],"labels.Moreover,":[141],"demonstrate":[143],"that":[144,164],"produced":[148],"student":[151],"benefit":[154],"training.Experiments":[157],"on":[158],"LJSpeech":[159],"LibriSpeech":[161],"corpora":[162],"show":[163],"compared":[165],"supervised":[167],"baselines,":[168],"our":[169],"achieves":[172],"12.2%":[174],"CER":[175],"improvement":[176],"in":[177,183],"single-speaker":[179],"setting":[180],"38.6%":[182],"multi-speaker":[185],"setting.":[186]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}