{"id":"https://openalex.org/W4406461749","doi":"https://doi.org/10.1109/slt61566.2024.10832258","title":"Investigating Disentanglement in a Phoneme-Level Speech Codec for Prosody Modeling","display_name":"Investigating Disentanglement in a Phoneme-Level Speech Codec for Prosody Modeling","publication_year":2024,"publication_date":"2024-12-02","ids":{"openalex":"https://openalex.org/W4406461749","doi":"https://doi.org/10.1109/slt61566.2024.10832258"},"language":"en","primary_location":{"id":"doi:10.1109/slt61566.2024.10832258","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832258","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE Spoken Language Technology Workshop (SLT)","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/A5114367806","display_name":"Sotirios Karapiperis","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Sotirios Karapiperis","raw_affiliation_strings":["Innoetics, Samsung Electronics,Greece"],"affiliations":[{"raw_affiliation_string":"Innoetics, Samsung Electronics,Greece","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052196585","display_name":"Nikolaos Ellinas","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nikolaos Ellinas","raw_affiliation_strings":["Innoetics, Samsung Electronics,Greece"],"affiliations":[{"raw_affiliation_string":"Innoetics, Samsung Electronics,Greece","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015840293","display_name":"Alexandra Vioni","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Alexandra Vioni","raw_affiliation_strings":["Innoetics, Samsung Electronics,Greece"],"affiliations":[{"raw_affiliation_string":"Innoetics, Samsung Electronics,Greece","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113307596","display_name":"Junkwang Oh","orcid":null},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Junkwang Oh","raw_affiliation_strings":["Mobile eXperience Business, Samsung Electronics,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Mobile eXperience Business, Samsung Electronics,Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041656238","display_name":"Gunu Jho","orcid":null},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Gunu Jho","raw_affiliation_strings":["Mobile eXperience Business, Samsung Electronics,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Mobile eXperience Business, Samsung Electronics,Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101882326","display_name":"In-Chul Hwang","orcid":"https://orcid.org/0000-0002-0045-9984"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Inchul Hwang","raw_affiliation_strings":["Mobile eXperience Business, Samsung Electronics,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Mobile eXperience Business, Samsung Electronics,Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5111393130","display_name":"Spyros Raptis","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Spyros Raptis","raw_affiliation_strings":["Innoetics, Samsung Electronics,Greece"],"affiliations":[{"raw_affiliation_string":"Innoetics, Samsung Electronics,Greece","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5114367806"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.7283,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.78530846,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"668","last_page":"674"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998000264167786,"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":0.9998000264167786,"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":0.9973000288009644,"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/T10403","display_name":"Phonetics and Phonology Research","score":0.9972000122070312,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/prosody","display_name":"Prosody","score":0.8116130232810974},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6874029636383057},{"id":"https://openalex.org/keywords/codec","display_name":"Codec","score":0.6673458218574524},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.662325918674469},{"id":"https://openalex.org/keywords/speech-synthesis","display_name":"Speech synthesis","score":0.4292915165424347},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.08324337005615234}],"concepts":[{"id":"https://openalex.org/C542774811","wikidata":"https://www.wikidata.org/wiki/Q10880526","display_name":"Prosody","level":2,"score":0.8116130232810974},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6874029636383057},{"id":"https://openalex.org/C161765866","wikidata":"https://www.wikidata.org/wiki/Q184748","display_name":"Codec","level":2,"score":0.6673458218574524},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.662325918674469},{"id":"https://openalex.org/C14999030","wikidata":"https://www.wikidata.org/wiki/Q16346","display_name":"Speech synthesis","level":2,"score":0.4292915165424347},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.08324337005615234}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/slt61566.2024.10832258","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832258","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE Spoken Language Technology Workshop (SLT)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.699999988079071,"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W2046056978","https://openalex.org/W2107740512","https://openalex.org/W2107860279","https://openalex.org/W2130086727","https://openalex.org/W2752796333","https://openalex.org/W2962788625","https://openalex.org/W2963799213","https://openalex.org/W2964138190","https://openalex.org/W3015440759","https://openalex.org/W3016021263","https://openalex.org/W3033411150","https://openalex.org/W3091928890","https://openalex.org/W3092028330","https://openalex.org/W3097777922","https://openalex.org/W3194172310","https://openalex.org/W3195787187","https://openalex.org/W3215615641","https://openalex.org/W4285483692","https://openalex.org/W4289383906","https://openalex.org/W4307323391","https://openalex.org/W4311415873","https://openalex.org/W4313679638","https://openalex.org/W4320451749","https://openalex.org/W4366460484","https://openalex.org/W4377010126","https://openalex.org/W4381786045","https://openalex.org/W4391020683","https://openalex.org/W6750489868","https://openalex.org/W6755300632","https://openalex.org/W6778823374","https://openalex.org/W6783867762","https://openalex.org/W6784545093","https://openalex.org/W6795807602","https://openalex.org/W6840040786","https://openalex.org/W6847363464","https://openalex.org/W6848735303","https://openalex.org/W6849600165","https://openalex.org/W6851724922","https://openalex.org/W6853188576","https://openalex.org/W6936129901"],"related_works":["https://openalex.org/W1927421023","https://openalex.org/W10581632","https://openalex.org/W3149582125","https://openalex.org/W157238252","https://openalex.org/W2169632867","https://openalex.org/W2465421051","https://openalex.org/W152045069","https://openalex.org/W652196294","https://openalex.org/W2368700418","https://openalex.org/W1965141925"],"abstract_inverted_index":{"Most":[0],"of":[1,26,56,60,79,106,130],"the":[2,24,52,57,70,75,80,118],"prevalent":[3],"approaches":[4],"in":[5,91],"speech":[6],"prosody":[7,53],"modeling":[8,54],"rely":[9],"on":[10,32,38,69,82,109],"learning":[11],"global":[12,88],"style":[13],"representations":[14,84,122],"or":[15],"a":[16,87,127],"continuous":[17],"latent":[18,121,142],"space":[19,59,143],"which":[20,35],"encode":[21],"and":[22,77,85,98,112,139,157],"transfer":[23],"attributes":[25],"reference":[27],"speech.":[28],"However,":[29],"recent":[30],"work":[31],"neural":[33],"codecs":[34],"are":[36],"based":[37,108],"Residual":[39],"Vector":[40],"Quantization":[41],"(RVQ)":[42],"already":[43],"shows":[44],"great":[45],"potential":[46],"offering":[47],"distinct":[48],"advantages.":[49],"We":[50,72,101],"investigate":[51],"capabilities":[55],"discrete":[58,120],"such":[61],"an":[62,103],"RVQ-VAE":[63],"model,":[64],"modifying":[65],"it":[66],"to":[67,93,115,146,155],"operate":[68],"phoneme-level.":[71],"condition":[73],"both":[74,96],"encoder":[76],"decoder":[78],"model":[81],"linguistic":[83],"apply":[86],"speaker":[89,99],"embedding":[90],"order":[92],"factor":[94],"out":[95,145],"phonetic":[97],"information.":[100],"conduct":[102],"extensive":[104],"set":[105],"investigations":[107],"subjective":[110],"experiments":[111],"objective":[113],"measures":[114],"show":[116],"that":[117,136],"phoneme-level":[119],"obtained":[123],"this":[124],"way":[125],"achieve":[126],"high":[128],"degree":[129],"disentanglement,":[131],"capturing":[132],"fine-grained":[133],"prosodic":[134],"information":[135],"is":[137],"robust":[138],"transferable.":[140],"The":[141],"turns":[144],"have":[147],"interpretable":[148],"structure":[149],"with":[150],"its":[151],"principal":[152],"components":[153],"corresponding":[154],"pitch":[156],"energy.":[158]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-03-09T08:58:05.943551","created_date":"2025-10-10T00:00:00"}