{"id":"https://openalex.org/W7163670511","doi":"https://doi.org/10.48550/arxiv.2606.05367","title":"Task-Vector Arithmetic for Emotional Expressivity Control in Language-Model-Based Text-to-Speech","display_name":"Task-Vector Arithmetic for Emotional Expressivity Control in Language-Model-Based Text-to-Speech","publication_year":2026,"publication_date":"2026-06-03","ids":{"openalex":"https://openalex.org/W7163670511","doi":"https://doi.org/10.48550/arxiv.2606.05367"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.05367","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.05367","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.05367","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137920897","display_name":"Daniel Oliveira de Brito","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"de Brito, Daniel Oliveira","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137944199","display_name":"Arnaldo Candido Junior","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Junior, Arnaldo Candido","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10403","display_name":"Phonetics and Phonology Research","score":0.27639999985694885,"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"}},"topics":[{"id":"https://openalex.org/T10403","display_name":"Phonetics and Phonology Research","score":0.27639999985694885,"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"}},{"id":"https://openalex.org/T10667","display_name":"Emotion and Mood Recognition","score":0.1891999989748001,"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"}},{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.11590000241994858,"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/prosody","display_name":"Prosody","score":0.6589999794960022},{"id":"https://openalex.org/keywords/codec","display_name":"Codec","score":0.4779999852180481},{"id":"https://openalex.org/keywords/intelligibility","display_name":"Intelligibility (philosophy)","score":0.4406999945640564},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.39910000562667847},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.35420000553131104},{"id":"https://openalex.org/keywords/congruence","display_name":"Congruence (geometry)","score":0.3368000090122223},{"id":"https://openalex.org/keywords/operand","display_name":"Operand","score":0.31700000166893005},{"id":"https://openalex.org/keywords/discrete-cosine-transform","display_name":"Discrete cosine transform","score":0.3154999911785126}],"concepts":[{"id":"https://openalex.org/C542774811","wikidata":"https://www.wikidata.org/wiki/Q10880526","display_name":"Prosody","level":2,"score":0.6589999794960022},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5586000084877014},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5507000088691711},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.5454000234603882},{"id":"https://openalex.org/C161765866","wikidata":"https://www.wikidata.org/wiki/Q184748","display_name":"Codec","level":2,"score":0.4779999852180481},{"id":"https://openalex.org/C60048801","wikidata":"https://www.wikidata.org/wiki/Q1433889","display_name":"Intelligibility (philosophy)","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.39910000562667847},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.35420000553131104},{"id":"https://openalex.org/C132074034","wikidata":"https://www.wikidata.org/wiki/Q154210","display_name":"Congruence (geometry)","level":2,"score":0.3368000090122223},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.328900009393692},{"id":"https://openalex.org/C55526617","wikidata":"https://www.wikidata.org/wiki/Q719375","display_name":"Operand","level":2,"score":0.31700000166893005},{"id":"https://openalex.org/C2221639","wikidata":"https://www.wikidata.org/wiki/Q2877","display_name":"Discrete cosine transform","level":3,"score":0.3154999911785126},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.31189998984336853},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.29829999804496765},{"id":"https://openalex.org/C146599234","wikidata":"https://www.wikidata.org/wiki/Q511093","display_name":"Centroid","level":2,"score":0.2928999960422516},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.29010000824928284},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.289000004529953},{"id":"https://openalex.org/C45273575","wikidata":"https://www.wikidata.org/wiki/Q578970","display_name":"Spectrogram","level":2,"score":0.2849999964237213},{"id":"https://openalex.org/C178009071","wikidata":"https://www.wikidata.org/wiki/Q93344","display_name":"Trigonometric functions","level":2,"score":0.2809999883174896},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.26579999923706055},{"id":"https://openalex.org/C2777200299","wikidata":"https://www.wikidata.org/wiki/Q52943","display_name":"Conversation","level":2,"score":0.2615000009536743},{"id":"https://openalex.org/C128107574","wikidata":"https://www.wikidata.org/wiki/Q182003","display_name":"Injective function","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.2572999894618988},{"id":"https://openalex.org/C177291462","wikidata":"https://www.wikidata.org/wiki/Q423038","display_name":"Active listening","level":2,"score":0.25130000710487366}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.05367","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.05367","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.05367","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.05367","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.5665715336799622,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"We":[0],"investigate":[1],"whether":[2],"task-vector":[3],"arithmetic,":[4],"successful":[5],"for":[6,162],"cross-speaker":[7],"emotional":[8,75,184],"intensity":[9],"control":[10],"in":[11,93,136,172,186],"modular":[12],"text-to-speech":[13],"(TTS),":[14],"transfers":[15],"to":[16,77,103,195],"large-scale":[17],"TTS":[18],"systems":[19],"built":[20],"on":[21,38,81,90,143,149],"language-model":[22],"backbones":[23],"with":[24,64],"in-context":[25],"learning":[26],"(LM-TTS).":[27],"Through":[28],"a":[29,86,126],"systematic":[30],"elimination":[31],"study":[32],"over":[33,139],"four":[34],"progressively":[35],"narrower":[36],"operands":[37],"Qwen3-TTS-12Hz-1.7B":[39],"-":[40,68],"model":[41],"weights":[42],"via":[43],"LoRA":[44],"fine-tuning,":[45],"continuous":[46],"codec":[47,50],"embeddings,":[48],"discrete":[49],"tokens,":[51],"and":[52,121,147,167],"the":[53,65,71,78,118,140,163,180,196],"speaker":[54,107,198],"embedding":[55],"(x-vector)":[56],"produced":[57],"by":[58,193],"an":[59,96,104],"ECAPA-TDNN":[60],"encoder":[61],"jointly":[62],"trained":[63],"synthesis":[66],"backbone":[67],"we":[69,84,130],"localize":[70],"dominant":[72,181],"carrier":[73,182],"of":[74,134,183,189],"prosody":[76,185],"x-vector.":[79],"Building":[80],"this":[82,187],"finding,":[83],"propose":[85],"training-free":[87,201],"method":[88],"based":[89],"centroid":[91,202],"arithmetic":[92,203],"x-vector":[94],"space:":[95],"emotion":[97],"direction":[98],"$\u03c4=":[99],"\\mathbb{E}_i[x(s_i,\\text{emo})]":[100],"-\\mathbb{E}_i[x(s_i,\\text{neutral})]$":[101],"applied":[102],"unseen":[105],"target":[106],"as":[108,117,125],"$x_{\\text{new}}":[109],"=":[110],"x(\\text{target},\\text{neutral})":[111],"+":[112],"\u03b1\\cdot\u03c4$.":[113],"Using":[114],"ESD":[115],"(English)":[116],"$\u03c4$":[119,165],"source":[120],"emoUERJ":[122],"(Brazilian":[123],"Portuguese)":[124],"cross-lingual":[127,208],"ground-truth":[128],"target,":[129],"observe":[131],"average":[132],"gains":[133],"$+0.29$":[135],"emotion2vec":[137],"cosine":[138],"ICL":[141],"baseline":[142],"English":[144],"held-out":[145,152],"speakers":[146],"$+0.09$":[148],"Brazilian":[150],"Portuguese":[151],"speakers,":[153],"while":[154],"largely":[155],"preserving":[156],"identity":[157],"(WavLM":[158],"SECS":[159],"$\\gtrsim":[160],"0.88$":[161],"multi-speaker":[164],"variant)":[166],"intelligibility":[168],"(WER":[169],"$\\approx":[170],"0$":[171],"PT-BR).":[173],"These":[174],"results":[175],"offer":[176],"initial":[177],"evidence":[178],"that":[179],"class":[188],"models":[190],"is":[191],"localizable,":[192],"elimination,":[194],"co-trained":[197],"embedding,":[199],"where":[200],"remains":[204],"effective":[205],"even":[206],"under":[207],"transfer.":[209]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-06T00:00:00"}