{"id":"https://openalex.org/W4412090093","doi":"https://doi.org/10.1007/s11227-025-07473-7","title":"Saffe: Multimodal Model Composition with Semantic-Alignment Fusion of Frozen Encoders","display_name":"Saffe: Multimodal Model Composition with Semantic-Alignment Fusion of Frozen Encoders","publication_year":2025,"publication_date":"2025-07-07","ids":{"openalex":"https://openalex.org/W4412090093","doi":"https://doi.org/10.1007/s11227-025-07473-7"},"language":"en","primary_location":{"id":"doi:10.1007/s11227-025-07473-7","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-025-07473-7","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-025-07473-7.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s11227-025-07473-7.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5118848479","display_name":"Maithri Kulasekara","orcid":null},"institutions":[{"id":"https://openalex.org/I197250369","display_name":"General Sir John Kotelawala Defence University","ror":"https://ror.org/04n37he08","country_code":"LK","type":"education","lineage":["https://openalex.org/I197250369"]},{"id":"https://openalex.org/I80180929","display_name":"Universidad de Murcia","ror":"https://ror.org/03p3aeb86","country_code":"ES","type":"education","lineage":["https://openalex.org/I80180929"]}],"countries":["ES","LK"],"is_corresponding":true,"raw_author_name":"Maithri Kulasekara","raw_affiliation_strings":["Computer Engineering Dept., General Sir John Kotelawala Defence, Rathmalana, Sri Lanka","Dept. of Computer Engineering and Technology, Universidad de Murcia, Murcia, 30100, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Computer Engineering Dept., General Sir John Kotelawala Defence, Rathmalana, Sri Lanka","institution_ids":["https://openalex.org/I197250369"]},{"raw_affiliation_string":"Dept. of Computer Engineering and Technology, Universidad de Murcia, Murcia, 30100, Spain","institution_ids":["https://openalex.org/I80180929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040878076","display_name":"Juan F. Ingl\u00e9s-Romero","orcid":"https://orcid.org/0000-0002-3648-0267"},"institutions":[{"id":"https://openalex.org/I80180929","display_name":"Universidad de Murcia","ror":"https://ror.org/03p3aeb86","country_code":"ES","type":"education","lineage":["https://openalex.org/I80180929"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Juan F. Ingl\u00e9s-Romero","raw_affiliation_strings":["Emerging Tech Dept., Libelium LAB, Urban Center, Murcia, 30006, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Emerging Tech Dept., Libelium LAB, Urban Center, Murcia, 30006, Spain","institution_ids":["https://openalex.org/I80180929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028232016","display_name":"Baldomero Imbern\u00f3n","orcid":"https://orcid.org/0000-0002-2758-8364"},"institutions":[{"id":"https://openalex.org/I68763199","display_name":"Universidad Cat\u00f3lica de Murcia","ror":"https://ror.org/05b1rsv17","country_code":"ES","type":"education","lineage":["https://openalex.org/I68763199"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Baldomero Imbern\u00f3n","raw_affiliation_strings":["Dept. of Computer Science, Universidad Cat\u00f3lica de Murcia, Murcia, 30107, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Computer Science, Universidad Cat\u00f3lica de Murcia, Murcia, 30107, Spain","institution_ids":["https://openalex.org/I68763199"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007832959","display_name":"Jos\u00e9 Luis Abell\u00e1n","orcid":"https://orcid.org/0000-0003-3550-720X"},"institutions":[{"id":"https://openalex.org/I80180929","display_name":"Universidad de Murcia","ror":"https://ror.org/03p3aeb86","country_code":"ES","type":"education","lineage":["https://openalex.org/I80180929"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Jos\u00e9 L. Abell\u00e1n","raw_affiliation_strings":["Dept. of Computer Engineering and Technology, Universidad de Murcia, Murcia, 30100, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Computer Engineering and Technology, Universidad de Murcia, Murcia, 30100, Spain","institution_ids":["https://openalex.org/I80180929"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5118848479"],"corresponding_institution_ids":["https://openalex.org/I197250369","https://openalex.org/I80180929"],"apc_list":{"value":2990,"currency":"USD","value_usd":2990},"apc_paid":{"value":2990,"currency":"USD","value_usd":2990},"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.06260553,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"81","issue":"10","first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9970999956130981,"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":0.9970999956130981,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9962999820709229,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.9954000115394592,"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":0.9047141075134277},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.6801203489303589},{"id":"https://openalex.org/keywords/composition","display_name":"Composition (language)","score":0.5859269499778748},{"id":"https://openalex.org/keywords/fusion","display_name":"Fusion","score":0.5806047320365906},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.47674593329429626},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4451347291469574},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.06843435764312744},{"id":"https://openalex.org/keywords/linguistics","display_name":"Linguistics","score":0.06266885995864868}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.9047141075134277},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6801203489303589},{"id":"https://openalex.org/C40231798","wikidata":"https://www.wikidata.org/wiki/Q1333743","display_name":"Composition (language)","level":2,"score":0.5859269499778748},{"id":"https://openalex.org/C158525013","wikidata":"https://www.wikidata.org/wiki/Q2593739","display_name":"Fusion","level":2,"score":0.5806047320365906},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.47674593329429626},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4451347291469574},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.06843435764312744},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.06266885995864868},{"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:10.1007/s11227-025-07473-7","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-025-07473-7","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-025-07473-7.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1007/s11227-025-07473-7","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-025-07473-7","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-025-07473-7.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3480869486","display_name":null,"funder_award_id":"13039","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G5080475149","display_name":null,"funder_award_id":"10.13039","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G7084143925","display_name":null,"funder_award_id":"AEI/10","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G7535663061","display_name":null,"funder_award_id":"AEI/10.","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320322805","display_name":"Universidad de Murcia","ror":"https://ror.org/03p3aeb86"},{"id":"https://openalex.org/F4320335598","display_name":"Agencia Estatal de Investigaci\u00f3n","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412090093.pdf","grobid_xml":"https://content.openalex.org/works/W4412090093.grobid-xml"},"referenced_works_count":44,"referenced_works":["https://openalex.org/W1861492603","https://openalex.org/W1981276685","https://openalex.org/W2108598243","https://openalex.org/W2742097923","https://openalex.org/W2766788829","https://openalex.org/W2883429621","https://openalex.org/W2963091558","https://openalex.org/W2970641574","https://openalex.org/W2990138404","https://openalex.org/W3035333188","https://openalex.org/W3096609285","https://openalex.org/W3134735773","https://openalex.org/W3143320354","https://openalex.org/W3148187679","https://openalex.org/W3176445421","https://openalex.org/W3196974791","https://openalex.org/W3201770677","https://openalex.org/W3202167223","https://openalex.org/W3204266507","https://openalex.org/W4200631575","https://openalex.org/W4206706211","https://openalex.org/W4214612132","https://openalex.org/W4312372834","https://openalex.org/W4312384316","https://openalex.org/W4313156423","https://openalex.org/W4386065512","https://openalex.org/W4386071707","https://openalex.org/W4386076403","https://openalex.org/W4387122290","https://openalex.org/W4389371459","https://openalex.org/W4390873312","https://openalex.org/W4393156194","https://openalex.org/W4394564148","https://openalex.org/W4402667090","https://openalex.org/W4402703116","https://openalex.org/W4402754197","https://openalex.org/W4402803808","https://openalex.org/W4406416302","https://openalex.org/W4406461926","https://openalex.org/W4407451021","https://openalex.org/W6600002382","https://openalex.org/W6600195515","https://openalex.org/W6600704668","https://openalex.org/W6636568255"],"related_works":["https://openalex.org/W4390516098","https://openalex.org/W2181948922","https://openalex.org/W2384362569","https://openalex.org/W2142795561","https://openalex.org/W1499958165","https://openalex.org/W4205302943","https://openalex.org/W2561132942","https://openalex.org/W2380193321","https://openalex.org/W3155418658","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Abstract":[0],"Transformer-based":[1],"multimodal":[2,35,131],"models":[3,36,128],"often":[4],"require":[5],"expensive,":[6],"full-model":[7],"training":[8,103,150],"on":[9,17,82,120,133],"task-specific":[10,34],"all-modality":[11],"datasets":[12,134],"to":[13,86],"achieve":[14],"high":[15],"accuracy":[16,107],"targeted":[18],"downstream":[19,110],"tasks.":[20,126],"To":[21],"reduce":[22],"this":[23],"significant":[24],"cost,":[25],"we":[26,117],"introduce":[27],"SAFFE,":[28],"a":[29,65,76,113],"methodology":[30],"for":[31,52,101,108],"building":[32],"accurate,":[33],"with":[37,144],"minimal":[38],"training,":[39],"using":[40],"only":[41],"standard":[42],"GPU":[43],"hardware.":[44],"SAFFE":[45,119],"leverages":[46],"off-the-shelf,":[47],"pre-trained,":[48],"frozen":[49],"unimodal":[50,94],"encoders":[51],"each":[53],"input":[54],"modality":[55],"(e.g.,":[56],"text,":[57],"image,":[58],"or":[59],"audio)":[60],"and":[61,123,139,149],"connects":[62],"them":[63],"through":[64],"lightweight,":[66],"trainable":[67,147],"component":[68],"called":[69],"the":[70,83,88,91,99,109],"FusionAlign":[71],"Module":[72],"(FAM).":[73],"FAM":[74],"is":[75],"bottleneck":[77],"mid-fusion":[78],"neural":[79],"network,":[80],"trained":[81],"target":[84],"dataset":[85],"align":[87],"outputs":[89],"of":[90,115],"independently":[92],"pre-trained":[93],"encoders.":[95],"This":[96],"approach":[97],"eliminates":[98],"need":[100],"end-to-end":[102],"while":[104],"maintaining":[105],"strong":[106],"task.":[111],"As":[112],"proof":[114],"concept,":[116],"validate":[118],"image":[121],"retrieval":[122],"language":[124],"understanding":[125],"SAFFE-derived":[127],"outperform":[129],"state-of-the-art":[130],"systems":[132],"such":[135],"as":[136],"CIFAR-10,":[137],"ImageNet-100,":[138],"COCO,":[140],"achieving":[141],"competitive":[142],"results":[143],"significantly":[145],"fewer":[146],"parameters":[148],"time.":[151]},"counts_by_year":[],"updated_date":"2026-07-23T08:03:31.855105","created_date":"2025-10-10T00:00:00"}