{"id":"https://openalex.org/W4415334308","doi":"https://doi.org/10.48550/arxiv.2506.15645","title":"Demystifying the Visual Quality Paradox in Multimodal Large Language Models","display_name":"Demystifying the Visual Quality Paradox in Multimodal Large Language Models","publication_year":2025,"publication_date":"2025-06-18","ids":{"openalex":"https://openalex.org/W4415334308","doi":"https://doi.org/10.48550/arxiv.2506.15645"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2506.15645","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.15645","pdf_url":"https://arxiv.org/pdf/2506.15645","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"parent_works":[],"lineage":["https://openalex.org/W4415334308"],"child_works_count":0,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2506.15645","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5108920201","display_name":"Shuo Xing","orcid":"https://orcid.org/0000-0002-5585-6215","observed_orcids":["https://orcid.org/0000-0002-5585-6215"]},"institutions":[{"id":"https://openalex.org/I91045830","display_name":"Texas A&M University","ror":"https://ror.org/01f5ytq51","country_code":"US","type":"education","lineage":["https://openalex.org/I91045830"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xing, Shuo","raw_affiliation_strings":["Texas A&M University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Texas A&M University","institution_ids":["https://openalex.org/I91045830"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Guo, Lanqing","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I86519309","display_name":"The University of Texas at Austin","ror":"https://ror.org/00hj54h04","country_code":"US","type":"education","lineage":["https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Guo, Lanqing","raw_affiliation_strings":["University of Texas at Austin"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Texas at Austin","institution_ids":["https://openalex.org/I86519309"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hua, Hongyuan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Hua, Hongyuan","raw_affiliation_strings":["University of Toronto"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Toronto","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100661015","display_name":"Seoyoung Lee","orcid":"https://orcid.org/0000-0003-2821-1208","observed_orcids":["https://orcid.org/0000-0003-2821-1208"]},"institutions":[{"id":"https://openalex.org/I86519309","display_name":"The University of Texas at Austin","ror":"https://ror.org/00hj54h04","country_code":"US","type":"education","lineage":["https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lee, Seoyoung","raw_affiliation_strings":["University of Texas at Austin"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Texas at Austin","institution_ids":["https://openalex.org/I86519309"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101694512","display_name":"Peiran Li","orcid":"https://orcid.org/0000-0003-0751-2191","observed_orcids":["https://orcid.org/0000-0003-0751-2191"]},"institutions":[{"id":"https://openalex.org/I91045830","display_name":"Texas A&M University","ror":"https://ror.org/01f5ytq51","country_code":"US","type":"education","lineage":["https://openalex.org/I91045830"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li, Peiran","raw_affiliation_strings":["Texas A&M University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Texas A&M University","institution_ids":["https://openalex.org/I91045830"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Wang, Yufei","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Wang, Yufei","raw_affiliation_strings":["Nanyang Technological University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanyang Technological University","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048522863","display_name":"Zhangyang Wang","orcid":"https://orcid.org/0000-0002-2050-5693","observed_orcids":["https://orcid.org/0000-0002-2050-5693"]},"institutions":[{"id":"https://openalex.org/I86519309","display_name":"The University of Texas at Austin","ror":"https://ror.org/00hj54h04","country_code":"US","type":"education","lineage":["https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wang, Zhangyang","raw_affiliation_strings":["University of Texas at Austin"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Texas at Austin","institution_ids":["https://openalex.org/I86519309"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5015173810","display_name":"Zhengzhong Tu","orcid":"https://orcid.org/0000-0002-7594-2292","observed_orcids":["https://orcid.org/0000-0002-7594-2292"]},"institutions":[{"id":"https://openalex.org/I91045830","display_name":"Texas A&M University","ror":"https://ror.org/01f5ytq51","country_code":"US","type":"education","lineage":["https://openalex.org/I91045830"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tu, Zhengzhong","raw_affiliation_strings":["Texas A&M University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Texas A&M University","institution_ids":["https://openalex.org/I91045830"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":5,"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":1.0,"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":1.0,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.0,"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":0.9769999980926514},{"id":"https://openalex.org/keywords/image-degradation","display_name":"image degradation","score":0.6840000152587891},{"id":"https://openalex.org/keywords/image-stylization","display_name":"image stylization","score":0.7229999899864197},{"id":"https://openalex.org/keywords/low-rank-adaptation","display_name":"low-rank adaptation","score":0.8759999871253967},{"id":"https://openalex.org/keywords/image-restoration","display_name":"image restoration","score":0.593999981880188}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7512999773025513},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.6654999852180481},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5626999735832214},{"id":"https://openalex.org/C2780719617","wikidata":"https://www.wikidata.org/wiki/Q1030752","display_name":"Salient","level":2,"score":0.5414000153541565},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.5343000292778015},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.48890000581741333},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.4878000020980835},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.4772999882698059},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.4287000000476837},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.42669999599456787},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4129999876022339},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.37790000438690186},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3612000048160553},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.3571000099182129},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.35519999265670776},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.34439998865127563},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3352999985218048},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.33320000767707825},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.3050000071525574},{"id":"https://openalex.org/C2780878386","wikidata":"https://www.wikidata.org/wiki/Q1659648","display_name":"Visual language","level":2,"score":0.30090001225471497},{"id":"https://openalex.org/C2781235140","wikidata":"https://www.wikidata.org/wiki/Q275131","display_name":"Scratch","level":2,"score":0.2904999852180481},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.28790000081062317},{"id":"https://openalex.org/C160086991","wikidata":"https://www.wikidata.org/wiki/Q5939193","display_name":"Human visual system model","level":3,"score":0.2824999988079071},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2799000144004822},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.27219998836517334},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.25949999690055847},{"id":"https://openalex.org/C5339829","wikidata":"https://www.wikidata.org/wiki/Q1425977","display_name":"Machine vision","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2506.15645","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.15645","pdf_url":"https://arxiv.org/pdf/2506.15645","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2506.15645","doi":"https://doi.org/10.48550/arxiv.2506.15645","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2506.15645","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2506.15645","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.15645","pdf_url":"https://arxiv.org/pdf/2506.15645","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415334308.pdf","grobid_xml":"https://content.openalex.org/works/W4415334308.grobid-xml"},"referenced_works_count":38,"referenced_works":["https://openalex.org/W1889081078","https://openalex.org/W2896457183","https://openalex.org/W2961540362","https://openalex.org/W4309953147","https://openalex.org/W4318718620","https://openalex.org/W4320343008","https://openalex.org/W4322718191","https://openalex.org/W4366330503","https://openalex.org/W4382142077","https://openalex.org/W4384918448","https://openalex.org/W4385416182","https://openalex.org/W4386185600","https://openalex.org/W4386185625","https://openalex.org/W4390041933","https://openalex.org/W4391988276","https://openalex.org/W4392678661","https://openalex.org/W4395474395","https://openalex.org/W4399695759","https://openalex.org/W4400601172","https://openalex.org/W4400601342","https://openalex.org/W4401660192","https://openalex.org/W4403853618","https://openalex.org/W4404820176","https://openalex.org/W4405434026","https://openalex.org/W4405655797","https://openalex.org/W4407794264","https://openalex.org/W4409917551","https://openalex.org/W4414850555","https://openalex.org/W4414937603","https://openalex.org/W4415085804","https://openalex.org/W4415133854","https://openalex.org/W4415329043","https://openalex.org/W4416037075","https://openalex.org/W4416070646","https://openalex.org/W4417307872","https://openalex.org/W6910735377","https://openalex.org/W7075692159","https://openalex.org/W7133185515"],"related_works":["https://openalex.org/W4392538638","https://openalex.org/W7140107836","https://openalex.org/W7171808912","https://openalex.org/W7114924848","https://openalex.org/W4402671098","https://openalex.org/W4404536559","https://openalex.org/W4393027664","https://openalex.org/W4403780546","https://openalex.org/W6892000976","https://openalex.org/W4387076386"],"abstract_inverted_index":{"Recent":[0],"Multimodal":[1],"Large":[2],"Language":[3],"Models":[4],"(MLLMs)":[5],"excel":[6],"on":[7],"benchmark":[8],"vision-language":[9,48],"tasks,":[10],"yet":[11],"little":[12],"is":[13],"known":[14],"about":[15],"how":[16],"input":[17,130],"visual":[18,169],"quality":[19,26],"shapes":[20],"their":[21],"responses.":[22],"Does":[23],"higher":[24],"perceptual":[25],"of":[27,47,188],"images":[28,74],"already":[29],"translate":[30],"to":[31,56,83,113],"better":[32],"MLLM":[33],"understanding?":[34],"We":[35],"conduct":[36],"the":[37,90,109,144,175,185,191],"first":[38],"systematic":[39],"study":[40],"spanning":[41],"leading":[42],"MLLMs":[43,146,172],"and":[44,53,67,117,147,173],"a":[45,62,104,133],"suite":[46],"benchmarks,":[49],"applying":[50],"controlled":[51],"degradations":[52],"stylistic":[54],"shifts":[55],"each":[57,129],"image.":[58],"Surprisingly,":[59],"we":[60,92],"uncover":[61],"visual-quality":[63],"paradox:":[64],"model,":[65],"task,":[66],"even":[68],"individual-instance":[69],"performance":[70],"can":[71],"improve":[72],"when":[73],"deviate":[75],"from":[76],"human-perceived":[77],"fidelity.":[78],"Off-the-shelf":[79],"restoration":[80],"pipelines":[81],"fail":[82],"reconcile":[84],"these":[85],"idiosyncratic":[86],"preferences.":[87,142],"To":[88],"close":[89],"gap,":[91],"introduce":[93],"Visual-Quality":[94],"Test-Time":[95],"Tuning":[96],"(VQ-TTT)-a":[97],"lightweight":[98],"adaptation":[99],"module":[100],"that:":[101],"(1)":[102],"inserts":[103],"learnable,":[105],"low-rank":[106],"kernel":[107],"before":[108],"frozen":[110],"vision":[111],"encoder":[112],"modulate":[114],"frequency":[115],"content;":[116],"(2)":[118],"fine-tunes":[119],"only":[120],"shallow":[121],"vision-encoder":[122],"layers":[123],"via":[124],"LoRA.":[125],"VQ-TTT":[126,150],"dynamically":[127],"adjusts":[128],"image":[131],"in":[132,184],"single":[134],"forward":[135],"pass,":[136],"aligning":[137],"it":[138],"with":[139,155],"task-specific":[140],"model":[141],"Across":[143],"evaluated":[145],"all":[148],"datasets,":[149],"lifts":[151],"significant":[152],"average":[153],"accuracy,":[154],"no":[156],"external":[157],"models,":[158],"cached":[159],"features,":[160],"or":[161],"extra":[162],"training":[163],"data.":[164],"These":[165],"findings":[166],"redefine":[167],"``better''":[168],"inputs":[170],"for":[171,177],"highlight":[174],"need":[176],"adaptive,":[178],"rather":[179],"than":[180],"universally":[181],"``clean'',":[182],"imagery,":[183],"new":[186],"era":[187],"AI":[189],"being":[190],"main":[192],"data":[193],"customer.":[194]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":2}],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2025-10-19T00:00:00"}