{"id":"https://openalex.org/W7204452005","doi":"https://doi.org/10.48550/arxiv.2608.25305","title":"MulVec: Fine-Grained Role-Aware Matching for Training-Free Zero-Shot Composed Image Retrieval","display_name":"MulVec: Fine-Grained Role-Aware Matching for Training-Free Zero-Shot Composed Image Retrieval","publication_year":2026,"publication_date":"2026-08-26","ids":{"openalex":"https://openalex.org/W7204452005","doi":"https://doi.org/10.48550/arxiv.2608.25305"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.25305","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.25305","pdf_url":"https://arxiv.org/pdf/2608.25305","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,"listed_in":[],"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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.25305","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148467849","display_name":"Zihao Zhang","orcid":"https://orcid.org/0000-0001-6859-7518","observed_orcids":["https://orcid.org/0000-0001-6859-7518"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Zihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002436544","display_name":"Dayan Wu","orcid":"https://orcid.org/0000-0002-8604-7226","observed_orcids":["https://orcid.org/0000-0002-8604-7226"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Dayan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023917479","display_name":"Xinze Liu","orcid":"https://orcid.org/0009-0006-4482-8995","observed_orcids":["https://orcid.org/0009-0006-4482-8995","https://orcid.org/0000-0002-3786-2646"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Xinze","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135142094","display_name":"Hengjie Zhu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Hengjie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5148425572","display_name":"Yiliang Zhu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Yiliang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133331301","display_name":"Ding Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Ding","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5119959312","display_name":"Peng Fu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5148437336","display_name":"Zheng Lin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Zheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129974097","display_name":"Weiping Wang","orcid":"https://orcid.org/0000-0002-8618-4992","observed_orcids":["https://orcid.org/0000-0002-8618-4992"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Weiping","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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.7534000277519226,"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"}},"topics":[{"id":"https://openalex.org/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.7534000277519226,"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/T10824","display_name":"Image Retrieval and Classification Techniques","score":0.17870000004768372,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.031199999153614044,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/zero-shot-image-retrieval","display_name":"zero-shot image retrieval","score":0.800000011920929},{"id":"https://openalex.org/keywords/composed-image-retrieval","display_name":"composed image retrieval","score":0.7839999794960022},{"id":"https://openalex.org/keywords/training-free-retrieval","display_name":"training-free retrieval","score":0.8320000171661377},{"id":"https://openalex.org/keywords/frozen-encoder","display_name":"frozen encoder","score":0.6890000104904175},{"id":"https://openalex.org/keywords/circos","display_name":"Circos","score":0.9879999756813049}],"concepts":[{"id":"https://openalex.org/C1667742","wikidata":"https://www.wikidata.org/wiki/Q10927554","display_name":"Image retrieval","level":3,"score":0.751800000667572},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7045000195503235},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6212999820709229},{"id":"https://openalex.org/C189391414","wikidata":"https://www.wikidata.org/wiki/Q7936579","display_name":"Visual Word","level":4,"score":0.5407999753952026},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5315999984741211},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.4413999915122986},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.4413999915122986},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.41589999198913574},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3912999927997589},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.3521000146865845},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.3499000072479248},{"id":"https://openalex.org/C199579030","wikidata":"https://www.wikidata.org/wiki/Q2851778","display_name":"Automatic image annotation","level":4,"score":0.3440000116825104},{"id":"https://openalex.org/C2986492983","wikidata":"https://www.wikidata.org/wiki/Q861092","display_name":"Image matching","level":3,"score":0.30169999599456787},{"id":"https://openalex.org/C2780052074","wikidata":"https://www.wikidata.org/wiki/Q1128648","display_name":"Content-based image retrieval","level":4,"score":0.2849000096321106},{"id":"https://openalex.org/C99016210","wikidata":"https://www.wikidata.org/wiki/Q5488129","display_name":"Query expansion","level":2,"score":0.2759000062942505},{"id":"https://openalex.org/C88516994","wikidata":"https://www.wikidata.org/wiki/Q1268863","display_name":"Dynamic time warping","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2513999938964844},{"id":"https://openalex.org/C158096908","wikidata":"https://www.wikidata.org/wiki/Q3983303","display_name":"Template matching","level":3,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.25305","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.25305","pdf_url":"https://arxiv.org/pdf/2608.25305","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,"listed_in":[],"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.2608.25305","doi":"https://doi.org/10.48550/arxiv.2608.25305","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.25305","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,"listed_in":[],"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":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.25305","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.25305","pdf_url":"https://arxiv.org/pdf/2608.25305","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,"listed_in":[],"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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7166855463","https://openalex.org/W7164826998","https://openalex.org/W7223067407","https://openalex.org/W7155414113","https://openalex.org/W7154749275","https://openalex.org/W4389820801","https://openalex.org/W7125897992","https://openalex.org/W7221309638","https://openalex.org/W7220666349","https://openalex.org/W4417119602"],"abstract_inverted_index":{"Training-free":[0],"zero-shot":[1],"composed":[2],"image":[3,8,15,24,42],"retrieval":[4,75,130,156],"finds":[5],"a":[6,10,13,17,33,40,60,66,123,142,154],"target":[7,31,105],"in":[9,153,188],"gallery":[11,152],"from":[12,22],"reference":[14],"and":[16,35,52,92,108,122,141,161,163,180,185],"text":[18],"edit":[19],"without":[20],"learning":[21],"task-specific":[23],"triplets.":[25],"Existing":[26],"methods":[27],"typically":[28],"describe":[29],"the":[30,79,101,150,176,182],"as":[32],"whole":[34],"match":[36],"this":[37,134],"description":[38,106],"with":[39],"global":[41,45,119],"representation.":[43],"This":[44],"matching":[46],"can":[47],"mix":[48],"different":[49],"semantic":[50],"cues":[51],"lose":[53],"fine-":[54],"grained":[55],"details.":[56],"We":[57],"propose":[58],"MULVEC,":[59],"role-aware":[61],"method":[62,179],"whose":[63],"compiler":[64],"produces":[65],"structured":[67],"query":[68,102],"record":[69],"that":[70],"is":[71,115],"mapped":[72],"to":[73,103,173],"four":[74],"roles:":[76],"Global":[77],"describes":[78],"full":[80],"target,":[81],"Desired":[82],"states":[83,88,94],"what":[84,89,95],"should":[85,90,96],"appear,":[86],"Preserve":[87],"remain,":[91],"Forbidden":[93],"disappear.":[97],"Frozen":[98],"encoders":[99],"map":[100],"one":[104,118],"vector":[107,121],"role-specific":[109],"probe":[110],"vectors,":[111],"while":[112],"each":[113],"candidate":[114],"represented":[116],"by":[117,171],"visual":[120,127],"bank":[124],"of":[125,146],"local":[126],"vectors.":[128],"The":[129],"roles":[131],"then":[132],"use":[133],"shared":[135],"evidence":[136],"for":[137],"their":[138,147],"respective":[139],"purposes,":[140],"fixed":[143],"weighted":[144],"sum":[145],"scores":[148],"ranks":[149],"entire":[151],"single":[155],"pass.":[157],"Across":[158],"CIRCO,":[159],"CIRR,":[160],"FashionIQ":[162,186],"three":[164],"backbone":[165],"scales,":[166],"MULVEC":[167],"improves":[168],"CIRCO":[169],"mAP@5":[170],"up":[172],"23.0%":[174],"over":[175],"strongest":[177],"compared":[178],"gives":[181],"best":[183],"CIRR":[184],"results":[187],"our":[189],"comparison.":[190]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-08-28T00:00:00"}