{"id":"https://openalex.org/W4404310810","doi":"https://doi.org/10.1109/iccv51701.2025.00030","title":"LLaVA-KD: A Framework of Distilling Multimodal Large Language Models","display_name":"LLaVA-KD: A Framework of Distilling Multimodal Large Language Models","publication_year":2025,"publication_date":"2025-10-19","ids":{"openalex":"https://openalex.org/W4404310810","doi":"https://doi.org/10.1109/iccv51701.2025.00030"},"language":"en","primary_location":{"id":"doi:10.1109/iccv51701.2025.00030","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccv51701.2025.00030","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF International Conference on Computer Vision (ICCV)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2410.16236","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101535508","display_name":"Yuxuan Cai","orcid":"https://orcid.org/0000-0003-0015-1653"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuxuan Cai","raw_affiliation_strings":["Huazhong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021861529","display_name":"Jiangning Zhang","orcid":"https://orcid.org/0000-0001-8891-6766"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiangning Zhang","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088106844","display_name":"Haoyang He","orcid":"https://orcid.org/0000-0001-6311-1611"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haoyang He","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101635302","display_name":"Xinwei He","orcid":"https://orcid.org/0000-0001-7267-0394"},"institutions":[{"id":"https://openalex.org/I204823248","display_name":"Huazhong Agricultural University","ror":"https://ror.org/023b72294","country_code":"CN","type":"education","lineage":["https://openalex.org/I204823248"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xinwei He","raw_affiliation_strings":["Huazhong Agricultural University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong Agricultural University","institution_ids":["https://openalex.org/I204823248"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114136681","display_name":"Ao Tong","orcid":null},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ao Tong","raw_affiliation_strings":["Huazhong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012172321","display_name":"Zhenye Gan","orcid":"https://orcid.org/0000-0003-1477-4958"},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenye Gan","raw_affiliation_strings":["Tencent Youtu Lab"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent Youtu Lab","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023834700","display_name":"Chengjie Wang","orcid":"https://orcid.org/0000-0003-4216-8090"},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chengjie Wang","raw_affiliation_strings":["Tencent Youtu Lab"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent Youtu Lab","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhucun Xue","orcid":null},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhucun Xue","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yong Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yong Liu","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"last","author":{"id":null,"display_name":"Xiang Bai","orcid":null},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiang Bai","raw_affiliation_strings":["Huazhong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology","institution_ids":["https://openalex.org/I47720641"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"239","last_page":"249"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9866999983787537,"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":0.9866999983787537,"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":0.9733999967575073,"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":0.5539803504943848},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.37937167286872864},{"id":"https://openalex.org/keywords/geology","display_name":"Geology","score":0.3432045876979828},{"id":"https://openalex.org/keywords/geochemistry","display_name":"Geochemistry","score":0.3425419330596924}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5539803504943848},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.37937167286872864},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.3432045876979828},{"id":"https://openalex.org/C17409809","wikidata":"https://www.wikidata.org/wiki/Q161764","display_name":"Geochemistry","level":1,"score":0.3425419330596924}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1109/iccv51701.2025.00030","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccv51701.2025.00030","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF International Conference on Computer Vision (ICCV)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2410.16236","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.16236","pdf_url":"https://arxiv.org/pdf/2410.16236","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2410.16236","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2410.16236","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2410.16236","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.16236","pdf_url":"https://arxiv.org/pdf/2410.16236","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G4189157326","display_name":null,"funder_award_id":"62225603","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4404310810.pdf"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2324615561","https://openalex.org/W2086120259","https://openalex.org/W2390279801","https://openalex.org/W2245170124","https://openalex.org/W2076393078","https://openalex.org/W4391913857","https://openalex.org/W2358668433"],"abstract_inverted_index":{"The":[0],"success":[1],"of":[2,11,20,32,119,174],"Large":[3,13],"Language":[4,14],"Models":[5,15],"(LLMs)":[6],"has":[7],"inspired":[8],"the":[9,25,117,120,129,141,163,172],"development":[10],"Multimodal":[12,78],"(MLLMs)":[16],"for":[17],"unified":[18],"understanding":[19,145],"vision":[21],"and":[22,29,90,93,147,168],"language.":[23],"However,":[24],"increasing":[26],"model":[27,164],"size":[28],"computational":[30,50],"complexity":[31],"large-scale":[33],"MLLMs":[34,44],"(l-MLLMs)":[35],"limit":[36],"their":[37],"use":[38],"in":[39,134],"resource-constrained":[40],"scenarios.":[41],"Although":[42],"small-scale":[43],"(s-MLLMs)":[45],"are":[46],"designed":[47],"to":[48,68,73,81,97,102,114,127,139,151],"reduce":[49],"costs,":[51],"they":[52],"typically":[53],"suffer":[54],"from":[55,71],"performance":[56,160],"degradation.":[57],"To":[58],"mitigate":[59],"this":[60],"limitation,":[61],"we":[62,76,108],"propose":[63,109],"a":[64,110],"novel":[65],"LLaVA-KD":[66],"framework":[67],"transfer":[69,82,98],"knowledge":[70],"l-MLLMs":[72],"s-MLLMs.":[74],"Specifically,":[75],"introduce":[77],"Distillation":[79,95],"(MDist)":[80],"teacher":[83,99],"model's":[84,100],"robust":[85],"representations":[86,133],"across":[87],"both":[88],"visual":[89,104],"linguistic":[91],"modalities,":[92],"Relation":[94],"(RDist)":[96],"ability":[101],"capture":[103],"token":[105],"relationships.":[106],"Additionally,":[107],"three-stage":[111],"training":[112],"scheme":[113],"fully":[115],"exploit":[116],"potential":[118],"proposed":[121,176],"distillation":[122],"strategy:":[123],"1)":[124],"Distilled":[125,149],"Pre-Training":[126],"strengthen":[128],"alignment":[130],"between":[131],"visual-linguistic":[132],"s-MLLMs,":[135],"2)":[136],"Supervised":[137],"Fine-Tuning":[138,150],"equip":[140],"s-MLLMs":[142,159],"with":[143],"multimodal":[144],"capacity,":[146],"3)":[148],"refine":[152],"s-MLLM's":[153],"knowledge.":[154],"Our":[155],"approach":[156],"significantly":[157],"improves":[158],"without":[161],"altering":[162],"architecture.":[165],"Extensive":[166],"experiments":[167],"ablation":[169],"studies":[170],"validate":[171],"effectiveness":[173],"each":[175],"component.":[177],"Code":[178],"will":[179],"be":[180],"available":[181],"at":[182],"https://github.com/Fantasyele/LLaVA-KD.":[183]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}