{"id":"https://openalex.org/W4385823505","doi":"https://doi.org/10.21437/interspeech.2023-1097","title":"PronScribe: Highly Accurate Multimodal Phonemic Transcription From Speech and Text","display_name":"PronScribe: Highly Accurate Multimodal Phonemic Transcription From Speech and Text","publication_year":2023,"publication_date":"2023-08-14","ids":{"openalex":"https://openalex.org/W4385823505","doi":"https://doi.org/10.21437/interspeech.2023-1097"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2023-1097","is_oa":true,"landing_page_url":"https://doi.org/10.21437/interspeech.2023-1097","pdf_url":"https://www.isca-archive.org/interspeech_2023/yu23_interspeech.pdf","source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"INTERSPEECH 2023","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.isca-archive.org/interspeech_2023/yu23_interspeech.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113405494","display_name":"Yang Yu","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":true,"raw_author_name":"Yang Yu","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037409489","display_name":"Matthew Perez","orcid":"https://orcid.org/0009-0005-5348-3020"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":false,"raw_author_name":"Matthew Perez","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024316712","display_name":"Ankur Bapna","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":false,"raw_author_name":"Ankur Bapna","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5092639662","display_name":"Fadi Haik","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":false,"raw_author_name":"Fadi Haik","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065411002","display_name":"Siamak Tazari","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":false,"raw_author_name":"Siamak Tazari","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100433691","display_name":"Yu Zhang","orcid":"https://orcid.org/0000-0003-1100-4835"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]},{"id":"https://openalex.org/I4210117425","display_name":"Google (Israel)","ror":"https://ror.org/02c20ys54","country_code":"IL","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210117425","https://openalex.org/I4210128969"]}],"countries":["IL","US"],"is_corresponding":false,"raw_author_name":"Yu Zhang","raw_affiliation_strings":["Google USA , University of Michigan  , Google Israel"],"affiliations":[{"raw_affiliation_string":"Google USA , University of Michigan  , Google Israel","institution_ids":["https://openalex.org/I4210117425","https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5113405494"],"corresponding_institution_ids":["https://openalex.org/I1291425158","https://openalex.org/I4210117425"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.09869263,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"5541","last_page":"5545"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.975600004196167,"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":0.975600004196167,"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/computer-science","display_name":"Computer science","score":0.7377254366874695},{"id":"https://openalex.org/keywords/transcription","display_name":"Transcription (linguistics)","score":0.629417896270752},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5924423336982727},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4846132695674896},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3616960644721985},{"id":"https://openalex.org/keywords/linguistics","display_name":"Linguistics","score":0.15015316009521484}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7377254366874695},{"id":"https://openalex.org/C179926584","wikidata":"https://www.wikidata.org/wiki/Q207714","display_name":"Transcription (linguistics)","level":2,"score":0.629417896270752},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5924423336982727},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4846132695674896},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3616960644721985},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.15015316009521484},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2023-1097","is_oa":true,"landing_page_url":"https://doi.org/10.21437/interspeech.2023-1097","pdf_url":"https://www.isca-archive.org/interspeech_2023/yu23_interspeech.pdf","source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"INTERSPEECH 2023","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.21437/interspeech.2023-1097","is_oa":true,"landing_page_url":"https://doi.org/10.21437/interspeech.2023-1097","pdf_url":"https://www.isca-archive.org/interspeech_2023/yu23_interspeech.pdf","source":null,"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"INTERSPEECH 2023","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.47999998927116394}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4385823505.pdf","grobid_xml":"https://content.openalex.org/works/W4385823505.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825","https://openalex.org/W2028097510","https://openalex.org/W2505877856"],"abstract_inverted_index":{"We":[0],"present":[1],"PronScribe,":[2],"a":[3,21,46,96],"novel":[4],"method":[5],"for":[6,123],"phonemic":[7],"transcription":[8],"from":[9,77,90],"speech":[10,93],"and":[11,18,92,99],"text":[12,91,107],"input":[13],"based":[14],"on":[15],"careful":[16],"finetuning":[17],"adaptation":[19],"of":[20,34,38,48,61],"massive,":[22],"multilingual,":[23],"multimodal":[24],"speech-text":[25],"pretrained":[26],"model.We":[27],"show":[28,69],"that":[29,70],"our":[30],"model":[31],"is":[32,64,100],"capable":[33],"phonemically":[35],"transcribing":[36],"pronunciations":[37],"full":[39],"utterances":[40],"with":[41],"accurate":[42],"word":[43],"boundaries":[44],"in":[45,58,86,95,116],"variety":[47],"languages":[49],"covering":[50],"diverse":[51],"phonological":[52],"phenomena,":[53],"achieving":[54],"phoneme":[55],"error":[56],"rates":[57],"the":[59,110],"vicinity":[60],"1-2%":[62],"which":[63],"comparable":[65],"to":[66],"human":[67],"transcribers.We":[68],"PronScribe":[71],"can":[72,119],"effectively":[73,120],"learn":[74],"this":[75],"task":[76],"relatively":[78],"little":[79],"training":[80],"data,":[81],"making":[82],"it":[83],"attractive":[84],"even":[85],"low-resource":[87],"settings.It":[88],"learns":[89],"simultaneously":[94],"coherent":[97],"way,":[98],"better":[101],"than":[102],"previous":[103],"models":[104],"using":[105],"speech,":[106],"or":[108],"both.Additionally,":[109],"model's":[111],"good":[112],"transfer":[113],"learning":[114],"characteristics":[115],"multilingual":[117],"settings":[118],"boost":[121],"performance":[122],"lower-resourced":[124],"languages.10.":[125]},"counts_by_year":[],"updated_date":"2026-03-14T06:41:57.775601","created_date":"2025-10-10T00:00:00"}
