{"id":"https://openalex.org/W4416033913","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.869","title":"Captioning for Text-Video Retrieval via Dual-Group Direct Preference Optimization","display_name":"Captioning for Text-Video Retrieval via Dual-Group Direct Preference Optimization","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416033913","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.869"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.869","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.869","pdf_url":"https://aclanthology.org/2025.findings-emnlp.869.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.869.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101482853","display_name":"Ji Soo Lee","orcid":"https://orcid.org/0009-0003-9055-5236"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji Soo Lee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107941574","display_name":"Byeonghak Ko","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Byungoh Ko","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024720667","display_name":"Jaewon Cho","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jaewon Cho","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Howoong Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Howoong Lee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Jaewoon Byun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jaewoon Byun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5084814930","display_name":"Hyunwoo J. Kim","orcid":"https://orcid.org/0000-0002-2181-9264"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hyunwoo J. Kim","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":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"16022","last_page":"16039"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.902999997138977,"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":0.902999997138977,"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/T11439","display_name":"Video Analysis and Summarization","score":0.062300000339746475,"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":0.006599999964237213,"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/closed-captioning","display_name":"Closed captioning","score":0.6899999976158142},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.48910000920295715},{"id":"https://openalex.org/keywords/term","display_name":"Term (time)","score":0.27549999952316284},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2624000012874603}],"concepts":[{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.6899999976158142},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6414999961853027},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5636000037193298},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.48910000920295715},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35530000925064087},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2827000021934509},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.27549999952316284},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2624000012874603},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2303999960422516},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.23000000417232513}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.869","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.869","pdf_url":"https://aclanthology.org/2025.findings-emnlp.869.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.869","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.869","pdf_url":"https://aclanthology.org/2025.findings-emnlp.869.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G7542886975","display_name":null,"funder_award_id":"NRF-2023R1A2C2005373","funder_id":"https://openalex.org/F4320322120","funder_display_name":"National Research Foundation of Korea"}],"funders":[{"id":"https://openalex.org/F4320320671","display_name":"National Research Foundation","ror":"https://ror.org/05s0g1g46"},{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"},{"id":"https://openalex.org/F4320328359","display_name":"Ministry of Science and ICT, South Korea","ror":"https://ror.org/01wpjm123"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416033913.pdf","grobid_xml":"https://content.openalex.org/works/W4416033913.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In":[0],"text-video":[1],"retrieval,":[2],"auxiliary":[3,147,166],"captions":[4,36,171],"are":[5,56,67],"often":[6],"used":[7],"to":[8,38,134,168],"enhance":[9],"video":[10,120],"understanding,":[11],"bridging":[12],"the":[13,16],"gap":[14],"between":[15,79,137],"modalities.While":[17],"recent":[18],"advances":[19],"in":[20],"multi-modal":[21],"large":[22],"language":[23,60],"models":[24],"(MLLMs)":[25],"have":[26],"enabled":[27],"strong":[28],"zeroshot":[29],"caption":[30,92,122,148],"generation,":[31],"we":[32,83,125,155],"observe":[33],"that":[34,74,89,110,131,157],"such":[35,63,144],"tend":[37],"be":[39],"generic":[40],"and":[41,121,149],"indistinguishable":[42],"across":[43,116],"visually":[44],"similar":[45],"videos,":[46],"limiting":[47],"their":[48],"utility":[49],"for":[50,71,172],"fine-grained":[51,170],"retrieval.Moreover,":[52],"conventional":[53],"captioning":[54,112],"approaches":[55],"typically":[57,69],"evaluated":[58],"using":[59,94],"generation":[61,93],"metrics,":[62],"as":[64,145],"BLEU,":[65],"which":[66],"not":[68],"tailored":[70],"retrieval":[72,87,95,129,161],"tasks":[73],"require":[75],"making":[76],"discriminative":[77],"distinctions":[78],"candidates.To":[80],"address":[81],"this,":[82],"propose":[84],"CaRe-DPO,":[85],"a":[86,106,150],"framework":[88],"directly":[90],"optimizes":[91],"relevance":[96],"scores.At":[97],"its":[98],"core":[99],"is":[100,174],"Dual-Group":[101],"Direct":[102],"Preference":[103],"Optimization":[104],"(DG-DPO),":[105],"novel":[107],"learning":[108],"strategy":[109],"supervises":[111],"by":[113,163],"modeling":[114],"preferences":[115],"groups":[117],"of":[118],"distinct":[119],"pairs.In":[123],"addition,":[124],"present":[126],"an":[127,146],"MLLM-based":[128],"model":[130],"incorporates":[132],"role-embeddings":[133],"better":[135],"distinguish":[136],"textual":[138],"inputs":[139],"with":[140],"different":[141],"functional":[142],"roles,":[143],"text":[151],"query.Through":[152],"extensive":[153],"experiments,":[154],"demonstrate":[156],"CaRe-DPO":[158],"significantly":[159],"enhances":[160],"performance":[162],"effectively":[164],"leveraging":[165],"knowledge":[167],"generate":[169],"retrieval.Code":[173],"available":[175],"at":[176],"https://github.com/mlvlab/CaReDPO.":[177]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-11-08T00:00:00"}