{"id":"https://openalex.org/W7164946429","doi":"https://doi.org/10.48550/arxiv.2606.16092","title":"VinQA: Visual Elements Interleaved Long-form Answer Generation for Real-World Multimodal Document QA","display_name":"VinQA: Visual Elements Interleaved Long-form Answer Generation for Real-World Multimodal Document QA","publication_year":2026,"publication_date":"2026-06-15","ids":{"openalex":"https://openalex.org/W7164946429","doi":"https://doi.org/10.48550/arxiv.2606.16092"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.16092","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.16092","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":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.16092","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111249436","display_name":"Young Rok Jang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jang, Young Rok","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042948850","display_name":"Hyesoo Kong","orcid":"https://orcid.org/0000-0002-3742-7433"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kong, Hyesoo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113151470","display_name":"Kyunghwan An","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"An, Kyunghwan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138747733","display_name":"Jae Sub Huh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huh, Jae Sub","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077582267","display_name":"Gyeonghun Kim","orcid":"https://orcid.org/0000-0003-2924-8387"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Gyeonghun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138740398","display_name":"Stanley Jungkyu Choi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Choi, Stanley Jungkyu","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9912999868392944,"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":0.9912999868392944,"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":0.0026000000070780516,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.00139999995008111,"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/modality","display_name":"Modality (human\u2013computer interaction)","score":0.579200029373169},{"id":"https://openalex.org/keywords/parsing","display_name":"Parsing","score":0.570900022983551},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.5479999780654907},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.392300009727478},{"id":"https://openalex.org/keywords/citation","display_name":"Citation","score":0.3517000079154968},{"id":"https://openalex.org/keywords/visual-approach","display_name":"Visual approach","score":0.3165999948978424},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.3158999979496002}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7617999911308289},{"id":"https://openalex.org/C2780226545","wikidata":"https://www.wikidata.org/wiki/Q6888030","display_name":"Modality (human\u2013computer interaction)","level":2,"score":0.579200029373169},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.570900022983551},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.5479999780654907},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4950999915599823},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.47909998893737793},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.4731000065803528},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.392300009727478},{"id":"https://openalex.org/C2778805511","wikidata":"https://www.wikidata.org/wiki/Q1713","display_name":"Citation","level":2,"score":0.3517000079154968},{"id":"https://openalex.org/C2777055276","wikidata":"https://www.wikidata.org/wiki/Q7936580","display_name":"Visual approach","level":2,"score":0.3165999948978424},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3158999979496002},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C2780878386","wikidata":"https://www.wikidata.org/wiki/Q1659648","display_name":"Visual language","level":2,"score":0.2874999940395355},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.27390000224113464},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.26750001311302185},{"id":"https://openalex.org/C146834321","wikidata":"https://www.wikidata.org/wiki/Q2979672","display_name":"Closure (psychology)","level":2,"score":0.26350000500679016},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.2508000135421753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.16092","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.16092","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":"doi:10.48550/arxiv.2606.16092","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.16092","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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Real-world":[0],"documents":[1,207],"combine":[2],"text":[3,53,116],"with":[4,50,79,92,208,257],"tables,":[5],"charts,":[6],"photographs,":[7],"and":[8,54,98,106,117,124,154,195,214],"diagrams":[9],"arranged":[10],"in":[11,56,237],"diverse":[12,215],"layouts,":[13],"yet":[14],"existing":[15],"research":[16],"on":[17,178,187,220],"multimodal":[18,139],"large":[19],"language":[20],"models":[21,171,186,249],"(MLLMs)":[22],"for":[23,39,69,205],"document":[24,58,72],"QA":[25],"predominantly":[26],"produces":[27],"text-only":[28],"responses,":[29],"underutilizing":[30],"these":[31,100,126],"visual":[32,45,96,119,165,212,251],"elements.":[33],"We":[34,156],"introduce":[35],"VinQA,":[36,221],"a":[37,138,226],"dataset":[38],"long-form":[40],"answer":[41,151],"generation":[42],"where":[43],"cited":[44],"elements":[46,97,128,252],"are":[47],"explicitly":[48],"interleaved":[49],"their":[51,80,193],"supporting":[52,259],"grounded":[55],"relevant":[57],"pages.":[59],"To":[60],"support":[61],"this":[62,197],"task,":[63],"we":[64,135],"study":[65],"two":[66],"encoding":[67],"methods":[68],"feeding":[70],"raw":[71],"page":[73,113],"images":[74,91],"into":[75],"an":[76,243],"MLLM,":[77],"along":[78,147],"visual-element":[81],"citation":[82,166,216],"mechanisms:":[83],"(1)":[84],"Page":[85,223],"Encoding,":[86,109],"which":[87,110],"directly":[88,163],"encodes":[89,121],"full-page":[90],"bounding":[93],"boxes":[94],"of":[95],"treats":[99],"boxed":[101],"regions":[102],"as":[103,129],"citable":[104,130],"units;":[105],"(2)":[107],"Modality":[108,199,238],"parses":[111],"each":[112],"to":[114,144,162],"extract":[115],"crop":[118],"elements,":[120,213],"them":[122],"separately,":[123],"uses":[125],"cropped":[127],"units.":[131],"In":[132],"our":[133],"experiments,":[134],"propose":[136],"M-GroSE,":[137],"evaluation":[140],"framework":[141],"extending":[142],"GroUSE":[143],"assess":[145],"answers":[146],"four":[148],"dimensions:":[149],"completeness,":[150],"relevancy,":[152],"faithfulness,":[153],"unanswerability.":[155],"additionally":[157],"report":[158],"Visual":[159,241],"Source":[160],"F1":[161],"measure":[164],"accuracy.":[167],"Although":[168],"proprietary":[169],"frontier":[170],"still":[172],"achieve":[173],"the":[174,179,188,233],"best":[175],"overall":[176],"scores":[177],"VinQA":[180],"test":[181],"split,":[182],"fine-tuning":[183],"open":[184],"Qwen2.5-VL":[185],"training":[189,219],"split":[190],"substantially":[191],"improves":[192],"performance":[194],"narrows":[196],"gap.":[198],"Encoding":[200,224],"is":[201],"initially":[202],"more":[203],"robust":[204],"complex":[206],"long":[209],"text,":[210],"many":[211],"requirements.":[217],"After":[218],"however,":[222],"reaches":[225],"comparable":[227],"level,":[228],"competing":[229],"effectively":[230],"even":[231],"without":[232],"explicit":[234],"parsing":[235],"used":[236],"Encoding.":[239],"Finally,":[240],"G-Eval,":[242],"MLLM-based":[244],"judge,":[245],"confirms":[246],"that":[247],"fine-tuned":[248],"insert":[250],"at":[253],"semantically":[254],"appropriate":[255],"positions":[256],"faithful":[258],"text.":[260]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-17T00:00:00"}