{"id":"https://openalex.org/W7077081376","doi":"https://doi.org/10.48550/arxiv.2508.03469","title":"IKOD: Mitigating Visual Attention Degradation in Large Vision-Language Models","display_name":"IKOD: Mitigating Visual Attention Degradation in Large Vision-Language Models","publication_year":2025,"publication_date":"2025-08-05","ids":{"openalex":"https://openalex.org/W7077081376","doi":"https://doi.org/10.48550/arxiv.2508.03469"},"language":"en","primary_location":{"id":"doi:10.48550/arxiv.2508.03469","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2508.03469","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2508.03469","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Yang, Jiabing","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yang, Jiabing","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Cui, Chenhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Chenhang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhou, Yiyang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Yiyang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Chen, Yixiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Yixiang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Xia, Peng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Peng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Wei, Ying","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei, Ying","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Yu, Tao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Tao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Huang, Yan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Yan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Wang, Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Liang","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"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":true,"primary_topic":{"id":"https://openalex.org/T13917","display_name":"Engineering and Material Science Research","score":0.061799999326467514,"subfield":{"id":"https://openalex.org/subfields/2500","display_name":"General Materials Science"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13917","display_name":"Engineering and Material Science Research","score":0.061799999326467514,"subfield":{"id":"https://openalex.org/subfields/2500","display_name":"General Materials Science"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T13370","display_name":"Diverse Scientific and Economic Studies","score":0.037700001150369644,"subfield":{"id":"https://openalex.org/subfields/2002","display_name":"Economics and Econometrics"},"field":{"id":"https://openalex.org/fields/20","display_name":"Economics, Econometrics and Finance"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12579","display_name":"Muon and positron interactions and applications","score":0.02070000022649765,"subfield":{"id":"https://openalex.org/subfields/2211","display_name":"Mechanics of Materials"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5719000101089478},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.5285000205039978},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.5005999803543091},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4966000020503998},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.46059998869895935},{"id":"https://openalex.org/keywords/face","display_name":"Face (sociological concept)","score":0.4334999918937683},{"id":"https://openalex.org/keywords/visual-attention","display_name":"Visual attention","score":0.3319000005722046}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6919000148773193},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5719000101089478},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5285000205039978},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.5005999803543091},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4966000020503998},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.46059998869895935},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45399999618530273},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.435699999332428},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.4334999918937683},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3517000079154968},{"id":"https://openalex.org/C2986089797","wikidata":"https://www.wikidata.org/wiki/Q6501338","display_name":"Visual attention","level":3,"score":0.3319000005722046},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.3174999952316284},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.3095000088214874},{"id":"https://openalex.org/C2779679103","wikidata":"https://www.wikidata.org/wiki/Q5251805","display_name":"Degradation (telecommunications)","level":2,"score":0.30149999260902405},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.28870001435279846},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.2678000032901764},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.26750001311302185},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.25769999623298645},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.2542000114917755}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2508.03469","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2508.03469","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2508.03469","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2508.03469","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":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0,69],"advancements":[1],"in":[2,40,100,121,219],"Large":[3],"Vision-Language":[4],"Models":[5],"(LVLMs)":[6],"have":[7,46],"demonstrated":[8],"significant":[9],"progress":[10],"across":[11],"multiple":[12],"domains.":[13],"However,":[14],"these":[15,51,151],"models":[16],"still":[17],"face":[18],"the":[19,41,82,87,104,124,132,190],"inherent":[20],"challenge":[21],"of":[22],"integrating":[23],"vision":[24],"and":[25,110,184,197,212,222,241],"language":[26],"for":[27,226],"collaborative":[28,163],"inference,":[29],"which":[30,136],"often":[31],"leads":[32],"to":[33,49,127,139,145,245],"\"hallucinations\",":[34],"outputs":[35],"that":[36,72],"are":[37],"not":[38,201],"grounded":[39],"corresponding":[42],"images.":[43],"Many":[44],"efforts":[45],"been":[47],"made":[48],"address":[50],"issues,":[52],"but":[53],"each":[54],"comes":[55],"with":[56,177,187],"its":[57],"own":[58],"limitations,":[59],"such":[60],"as":[61,81,131],"high":[62],"computational":[63],"cost":[64],"or":[65,234],"expensive":[66],"dataset":[67],"annotation.":[68],"research":[70,96],"shows":[71],"LVLMs":[73],"exhibit":[74],"a":[75,118,141,162,239],"long-term":[76,108],"bias":[77,109],"where":[78],"hallucinations":[79,199,221],"increase":[80],"sequence":[83,134],"length":[84],"grows,":[85,135],"yet":[86],"underlying":[88],"cause":[89],"remains":[90],"poorly":[91],"understood.":[92],"Building":[93],"on":[94,150,209],"extensive":[95],"into":[97],"attention":[98,126,180,195],"mechanisms":[99],"LVLMs,":[101],"we":[102,116,137,153],"analyze":[103],"relationship":[105],"between":[106],"this":[107],"visual":[111,128],"attention.":[112],"In":[113],"our":[114],"research,":[115],"identify":[117],"consistent":[119],"phenomenon":[120],"current":[122],"LVLMs:":[123],"model's":[125],"input":[129],"diminishes":[130],"generated":[133],"hypothesize":[138],"be":[140],"key":[142],"factor":[143],"contributing":[144],"observed":[146],"increasing":[147],"hallucinations.":[148],"Based":[149],"insights,":[152],"propose":[154],"Image":[155],"attention-guided":[156],"Key-value":[157],"merging":[158,183],"cOllaborative":[159],"Decoding":[160],"(IKOD),":[161],"decoding":[164],"strategy":[165],"generating":[166],"more":[167],"image-focused":[168],"sequences.":[169],"This":[170],"method":[171],"derives":[172],"logits":[173],"from":[174,189],"shorter":[175],"sequences":[176],"higher":[178],"image":[179],"through":[181],"key-value":[182],"combines":[185],"them":[186],"those":[188],"original":[191],"decoding,":[192],"effectively":[193],"mitigating":[194,220],"degradation":[196],"suppressing":[198],"while":[200],"incurring":[202],"too":[203],"much":[204],"inference":[205],"cost.":[206],"Extensive":[207],"experiments":[208],"both":[210],"hallucination":[211],"comprehensive":[213,224],"benchmarks":[214],"demonstrate":[215],"IKOD's":[216],"superior":[217],"effectiveness":[218],"improving":[223],"capacities":[225],"LVLMs.":[227],"Importantly,":[228],"IKOD":[229],"requires":[230],"no":[231],"additional":[232],"training":[233],"external":[235],"tools,":[236],"making":[237],"it":[238],"lightweight":[240],"efficient":[242],"framework":[243],"applicable":[244],"various":[246],"models.":[247]},"counts_by_year":[],"updated_date":"2025-11-06T06:51:31.235846","created_date":"2025-10-10T00:00:00"}