{"id":"https://openalex.org/W4408355438","doi":"https://doi.org/10.1109/icassp49660.2025.10890312","title":"EMMeTT: Efficient Multimodal Machine Translation Training","display_name":"EMMeTT: Efficient Multimodal Machine Translation Training","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408355438","doi":"https://doi.org/10.1109/icassp49660.2025.10890312"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10890312","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890312","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5027217976","display_name":"Piotr \u017belasko","orcid":"https://orcid.org/0000-0002-8245-0413"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Piotr \u017belasko","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002433660","display_name":"Zhehuai Chen","orcid":"https://orcid.org/0000-0003-4400-5340"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhehuai Chen","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051267437","display_name":"Mengru Wang","orcid":"https://orcid.org/0000-0002-4488-9871"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mengru Wang","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007998626","display_name":"Daniel G\u00e1lvez","orcid":"https://orcid.org/0000-0001-6119-4310"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Daniel Galvez","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088392468","display_name":"Oleksii Hrinchuk","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Oleksii Hrinchuk","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034588377","display_name":"Shuoyang Ding","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shuoyang Ding","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029338576","display_name":"Ke Hu","orcid":"https://orcid.org/0000-0002-1599-1519"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ke Hu","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040747392","display_name":"Jagadeesh Balam","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jagadeesh Balam","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026088310","display_name":"Vitaly Lavrukhin","orcid":"https://orcid.org/0009-0006-7866-8301"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Vitaly Lavrukhin","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5032957280","display_name":"Boris Ginsburg","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Boris Ginsburg","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5027217976"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":7.2362,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.96211663,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9983999729156494,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9983999729156494,"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/T10028","display_name":"Topic Modeling","score":0.9857000112533569,"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/T12031","display_name":"Speech and dialogue systems","score":0.9161999821662903,"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":0.7517682909965515},{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.6978064775466919},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.6817764043807983},{"id":"https://openalex.org/keywords/translation","display_name":"Translation (biology)","score":0.6066260933876038},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4751812517642975},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.3388960659503937}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7517682909965515},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.6978064775466919},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6817764043807983},{"id":"https://openalex.org/C149364088","wikidata":"https://www.wikidata.org/wiki/Q185917","display_name":"Translation (biology)","level":4,"score":0.6066260933876038},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4751812517642975},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3388960659503937},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C105580179","wikidata":"https://www.wikidata.org/wiki/Q188928","display_name":"Messenger RNA","level":3,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10890312","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890312","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W2327501763","https://openalex.org/W2963532001","https://openalex.org/W3001434439","https://openalex.org/W3034999214","https://openalex.org/W3082675885","https://openalex.org/W3169369929","https://openalex.org/W3208887030","https://openalex.org/W4223622550","https://openalex.org/W4319862635","https://openalex.org/W4385571651","https://openalex.org/W4389600306","https://openalex.org/W4392903872","https://openalex.org/W4402111558","https://openalex.org/W4402669706","https://openalex.org/W4406461487","https://openalex.org/W6631190155","https://openalex.org/W6767440493","https://openalex.org/W6767671539","https://openalex.org/W6769627184","https://openalex.org/W6778883912","https://openalex.org/W6789693907","https://openalex.org/W6810259195","https://openalex.org/W6810334672","https://openalex.org/W6850218400","https://openalex.org/W6855963333","https://openalex.org/W6856589991","https://openalex.org/W6856800273","https://openalex.org/W6857054612","https://openalex.org/W6859863615","https://openalex.org/W6860549370","https://openalex.org/W6861859833"],"related_works":["https://openalex.org/W2775554247","https://openalex.org/W2883671469","https://openalex.org/W2728761353","https://openalex.org/W2110168585","https://openalex.org/W3107474891","https://openalex.org/W2250213760","https://openalex.org/W4386247111","https://openalex.org/W4327642362","https://openalex.org/W2587014613","https://openalex.org/W3204019825"],"abstract_inverted_index":{"A":[0],"rising":[1],"interest":[2],"in":[3],"the":[4,14,80,126,154],"modality":[5],"extension":[6],"of":[7,37,137],"foundation":[8,49],"language":[9],"models":[10],"warrants":[11],"discussion":[12],"on":[13,25,134],"most":[15],"effective,":[16],"and":[17,30,54,87,93,139,149],"efficient,":[18],"multimodal":[19,34,65,100,113],"training":[20,35,71,77,114],"approach.":[21],"This":[22],"work":[23],"focuses":[24],"neural":[26],"machine":[27],"translation":[28,43,151],"(NMT)":[29],"proposes":[31],"a":[32,69,94,104,112],"joint":[33,64],"regime":[36],"Speech-LLM":[38],"to":[39],"include":[40],"automatic":[41],"speech":[42,60,150],"(AST).":[44],"We":[45,109],"investigate":[46],"two":[47],"different":[48],"model":[50],"architectures,":[51],"decoder-only":[52],"GPT":[53],"encoder-decoder":[55],"T5,":[56],"extended":[57],"with":[58,79,117,123],"Canary-1B\u2019s":[59],"encoder.":[61],"To":[62],"handle":[63],"training,":[66],"we":[67],"propose":[68],"novel":[70,95],"framework":[72],"called":[73],"EMMeTT.":[74],"EMMeTT":[75,124],"improves":[76],"efficiency":[78],"following:":[81],"balanced":[82],"sampling":[83],"across":[84],"languages,":[85],"datasets,":[86],"modalities;":[88],"efficient":[89],"sequential":[90],"data":[91],"iteration;":[92],"2D":[96],"bucketing":[97],"scheme":[98],"for":[99],"data,":[101],"complemented":[102],"by":[103],"batch":[105],"size":[106],"optimizer":[107],"(OOMptimizer).":[108],"show":[110],"that":[111],"consistently":[115],"helps":[116],"both":[118],"architectures.":[119],"Moreover,":[120],"SALM-T5":[121],"trained":[122],"retains":[125],"original":[127],"NMT":[128],"capability":[129],"while":[130],"outperforming":[131],"AST":[132],"baselines":[133],"four-language":[135],"subsets":[136],"FLORES":[138],"FLEURS.":[140],"The":[141],"resultant":[142],"Multimodal":[143],"Translation":[144],"Model":[145],"produces":[146],"strong":[147],"text":[148],"results":[152],"at":[153],"same":[155],"time.":[156]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2}],"updated_date":"2026-04-03T22:45:19.894376","created_date":"2025-10-10T00:00:00"}