{"id":"https://openalex.org/W4412376973","doi":"https://doi.org/10.1145/3726302.3730167","title":"Augmenting Vision-Language Retrieval: The Role of Multimodal LLMs as Synthetic Data Generators","display_name":"Augmenting Vision-Language Retrieval: The Role of Multimodal LLMs as Synthetic Data Generators","publication_year":2025,"publication_date":"2025-07-13","ids":{"openalex":"https://openalex.org/W4412376973","doi":"https://doi.org/10.1145/3726302.3730167"},"language":"en","primary_location":{"id":"doi:10.1145/3726302.3730167","doi":"https://doi.org/10.1145/3726302.3730167","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3726302.3730167","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3726302.3730167","source":{"id":"https://openalex.org/S4363608773","display_name":"International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3726302.3730167","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Aidan Bell","orcid":"https://orcid.org/0009-0001-1711-3189","observed_orcids":["https://orcid.org/0009-0001-1711-3189"]},"institutions":[{"id":"https://openalex.org/I73443619","display_name":"University of Southern Maine","ror":"https://ror.org/03ke6tv85","country_code":"US","type":"education","lineage":["https://openalex.org/I2802397601","https://openalex.org/I73443619"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Aidan Bell","raw_affiliation_strings":["University of Southern Maine, Portland, ME, USA"],"raw_orcid":"https://orcid.org/0009-0001-1711-3189","affiliations":[{"raw_affiliation_string":"University of Southern Maine, Portland, ME, USA","institution_ids":["https://openalex.org/I73443619"]}]},{"author_position":"middle","author":{"id":null,"display_name":"James Gore","orcid":"https://orcid.org/0009-0007-3771-855X","observed_orcids":["https://orcid.org/0009-0007-3771-855X"]},"institutions":[{"id":"https://openalex.org/I73443619","display_name":"University of Southern Maine","ror":"https://ror.org/03ke6tv85","country_code":"US","type":"education","lineage":["https://openalex.org/I2802397601","https://openalex.org/I73443619"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"James Gore","raw_affiliation_strings":["University of Southern Maine, Portland, ME, USA"],"raw_orcid":"https://orcid.org/0009-0007-3771-855X","affiliations":[{"raw_affiliation_string":"University of Southern Maine, Portland, ME, USA","institution_ids":["https://openalex.org/I73443619"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5078641935","display_name":"Behrooz Mansouri","orcid":"https://orcid.org/0000-0002-0400-9761","observed_orcids":["https://orcid.org/0000-0002-0400-9761"]},"institutions":[{"id":"https://openalex.org/I73443619","display_name":"University of Southern Maine","ror":"https://ror.org/03ke6tv85","country_code":"US","type":"education","lineage":["https://openalex.org/I2802397601","https://openalex.org/I73443619"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Behrooz Mansouri","raw_affiliation_strings":["University of Southern Maine, Portland, ME, USA"],"raw_orcid":"https://orcid.org/0000-0002-0400-9761","affiliations":[{"raw_affiliation_string":"University of Southern Maine, Portland, ME, USA","institution_ids":["https://openalex.org/I73443619"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I73443619"],"apc_list":null,"apc_paid":null,"fwci":1.5696,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.84557409,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"3050","last_page":"3054"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9883000254631042,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9883000254631042,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.008200000040233135,"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":0.002300000051036477,"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/multimodal-large-language-models","display_name":"multimodal large language models","score":0.8809999823570251},{"id":"https://openalex.org/keywords/synthetic-data-generation","display_name":"synthetic data generation","score":0.9139999747276306},{"id":"https://openalex.org/keywords/cross-modal-retrieval","display_name":"cross-modal retrieval","score":0.703000009059906},{"id":"https://openalex.org/keywords/caption-generation","display_name":"caption generation","score":0.6520000100135803},{"id":"https://openalex.org/keywords/blip","display_name":"BLIP","score":0.9990000128746033},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"fine-tuning","score":0.9210000038146973}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6238307952880859},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.45886537432670593},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4377460479736328}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3726302.3730167","doi":"https://doi.org/10.1145/3726302.3730167","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3726302.3730167","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3726302.3730167","source":{"id":"https://openalex.org/S4363608773","display_name":"International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3726302.3730167","doi":"https://doi.org/10.1145/3726302.3730167","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3726302.3730167","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3726302.3730167","source":{"id":"https://openalex.org/S4363608773","display_name":"International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412376973.pdf","grobid_xml":"https://content.openalex.org/works/W4412376973.grobid-xml"},"referenced_works_count":7,"referenced_works":["https://openalex.org/W2133512280","https://openalex.org/W3128777808","https://openalex.org/W3185239128","https://openalex.org/W3193658420","https://openalex.org/W4366483094","https://openalex.org/W6600669965","https://openalex.org/W6818723395"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Multimodal":[0],"Large":[1],"Language":[2],"Models":[3],"(MLLMs)":[4],"connect":[5],"and":[6,46,68,82],"interpret":[7],"different":[8],"data":[9],"types,":[10],"making":[11],"them":[12],"suitable":[13],"for":[14,26,71],"various":[15],"vision-language":[16],"tasks.":[17],"Despite":[18],"the":[19,40,58],"rapid":[20],"advancements":[21],"in":[22],"MLLMs,":[23],"their":[24],"effectiveness":[25,97],"specialized":[27],"cross-modal":[28,62,78],"retrieval":[29,63,79],"tasks":[30],"remains":[31],"underexplored.":[32],"A":[33],"challenging":[34],"example":[35],"is":[36,42],"art":[37],"retrieval,":[38],"where":[39],"task":[41],"to":[43,51,99],"find":[44],"visually":[45],"conceptually":[47],"relevant":[48],"artwork":[49],"corresponding":[50],"a":[52],"textual":[53],"description.":[54],"This":[55],"paper":[56],"investigates":[57],"effects":[59],"of":[60],"fine-tuning":[61],"models":[64,90],"using":[65],"both":[66],"human-annotated":[67,103],"MLLM-generated":[69,93],"captions":[70,94],"artistic":[72],"paintings.":[73],"To":[74],"this":[75],"end,":[76],"two":[77],"models,":[80],"Long-CLIP":[81],"BLIP,":[83],"are":[84],"studied.":[85],"Experimental":[86],"results":[87],"show":[88],"that":[89],"fine-tuned":[91,101],"on":[92,102],"achieve":[95],"search":[96],"comparable":[98],"those":[100],"captions.":[104]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2025-10-10T00:00:00"}