{"id":"https://openalex.org/W4415540430","doi":"https://doi.org/10.1145/3746027.3754784","title":"Identify, Isolate, and Purge: Mitigating Hallucinations in LVLMs via Self-Evolving Distillation","display_name":"Identify, Isolate, and Purge: Mitigating Hallucinations in LVLMs via Self-Evolving Distillation","publication_year":2025,"publication_date":"2025-10-25","ids":{"openalex":"https://openalex.org/W4415540430","doi":"https://doi.org/10.1145/3746027.3754784"},"language":"en","primary_location":{"id":"doi:10.1145/3746027.3754784","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3746027.3754784","pdf_url":null,"source":null,"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 33rd ACM International Conference on Multimedia","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3746027.3754784","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053573997","display_name":"Wenhao Li","orcid":"https://orcid.org/0009-0001-4010-3536"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]}],"countries":["AU"],"is_corresponding":true,"raw_author_name":"Wenhao Li","raw_affiliation_strings":["The University of Sydney, Sydney, Australia"],"raw_orcid":"https://orcid.org/0009-0001-4010-3536","affiliations":[{"raw_affiliation_string":"The University of Sydney, Sydney, Australia","institution_ids":["https://openalex.org/I129604602"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011334334","display_name":"Xiu Su","orcid":"https://orcid.org/0000-0002-9863-5404"},"institutions":[{"id":"https://openalex.org/I139660479","display_name":"Central South University","ror":"https://ror.org/00f1zfq44","country_code":"CN","type":"education","lineage":["https://openalex.org/I139660479"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiu Su","raw_affiliation_strings":["Central South University, Changsha, China"],"raw_orcid":"https://orcid.org/0000-0002-9863-5404","affiliations":[{"raw_affiliation_string":"Central South University, Changsha, China","institution_ids":["https://openalex.org/I139660479"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074317404","display_name":"Jingyi Wu","orcid":"https://orcid.org/0000-0001-8283-0646"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jingyi Wu","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0001-8283-0646","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002755427","display_name":"Feng Yang","orcid":"https://orcid.org/0000-0002-3495-5257"},"institutions":[{"id":"https://openalex.org/I76569877","display_name":"Southeast University","ror":"https://ror.org/04ct4d772","country_code":"CN","type":"education","lineage":["https://openalex.org/I76569877"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Feng Yang","raw_affiliation_strings":["Southeast University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-3495-5257","affiliations":[{"raw_affiliation_string":"Southeast University, Nanjing, China","institution_ids":["https://openalex.org/I76569877"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100355803","display_name":"Yang Liu","orcid":"https://orcid.org/0000-0002-1312-0146"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yang Liu","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-1312-0146","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103285702","display_name":"Yi Chen","orcid":"https://orcid.org/0000-0003-4283-7485"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Yi Chen","raw_affiliation_strings":["HKUST, Hong Kong, China"],"raw_orcid":"https://orcid.org/0000-0003-4283-7485","affiliations":[{"raw_affiliation_string":"HKUST, Hong Kong, China","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083588442","display_name":"Shan You","orcid":"https://orcid.org/0000-0003-1964-0430"},"institutions":[{"id":"https://openalex.org/I4210128910","display_name":"Group Sense (China)","ror":"https://ror.org/036wd5777","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210128910"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shan You","raw_affiliation_strings":["Sensetime Research, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1964-0430","affiliations":[{"raw_affiliation_string":"Sensetime Research, Beijing, China","institution_ids":["https://openalex.org/I4210128910"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001529504","display_name":"Chang Xu","orcid":"https://orcid.org/0000-0002-4756-0609"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Chang Xu","raw_affiliation_strings":["The University of Sydney, Sydney, Australia"],"raw_orcid":"https://orcid.org/0000-0002-4756-0609","affiliations":[{"raw_affiliation_string":"The University of Sydney, Sydney, Australia","institution_ids":["https://openalex.org/I129604602"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5053573997"],"corresponding_institution_ids":["https://openalex.org/I129604602"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.15621405,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"6791","last_page":"6800"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9287999868392944,"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"}},"topics":[{"id":"https://openalex.org/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9287999868392944,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6808000206947327},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.6075999736785889},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.4830000102519989},{"id":"https://openalex.org/keywords/credibility","display_name":"Credibility","score":0.4253000020980835},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.36079999804496765},{"id":"https://openalex.org/keywords/limit","display_name":"Limit (mathematics)","score":0.3458999991416931}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6808000206947327},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.6075999736785889},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5688999891281128},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.536300003528595},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5013999938964844},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.4830000102519989},{"id":"https://openalex.org/C2780224610","wikidata":"https://www.wikidata.org/wiki/Q1530061","display_name":"Credibility","level":2,"score":0.4253000020980835},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.36079999804496765},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.3458999991416931},{"id":"https://openalex.org/C2780704645","wikidata":"https://www.wikidata.org/wiki/Q9251458","display_name":"Observer (physics)","level":2,"score":0.30399999022483826},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2766999900341034},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.27489998936653137},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.2676999866962433},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.25839999318122864}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3746027.3754784","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3746027.3754784","pdf_url":null,"source":null,"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 33rd ACM International Conference on Multimedia","raw_type":"proceedings-article"},{"id":"pmh:oai:repository.hkust.edu.hk:1783.1-167660","is_oa":false,"landing_page_url":"http://repository.hkust.edu.hk/ir/Record/1783.1-167660","pdf_url":null,"source":{"id":"https://openalex.org/S4306401796","display_name":"Rare & Special e-Zone (The Hong Kong University of Science and Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I200769079","host_organization_name":"Hong Kong University of Science and Technology","host_organization_lineage":["https://openalex.org/I200769079"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Conference paper"}],"best_oa_location":{"id":"doi:10.1145/3746027.3754784","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3746027.3754784","pdf_url":null,"source":null,"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 33rd ACM International Conference on Multimedia","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":7,"referenced_works":["https://openalex.org/W2981851019","https://openalex.org/W3202965415","https://openalex.org/W4384662964","https://openalex.org/W4402704633","https://openalex.org/W4402727764","https://openalex.org/W4402753774","https://openalex.org/W4405595839"],"related_works":[],"abstract_inverted_index":{"Large":[0],"Vision-Language":[1],"Models":[2],"(LVLMs)":[3],"have":[4],"demonstrated":[5],"remarkable":[6],"advancements":[7],"in":[8,89,163],"numerous":[9],"areas":[10],"such":[11,170],"as":[12,171],"multimedia.":[13],"However,":[14],"hallucination":[15,183],"issues":[16],"significantly":[17,39],"limit":[18],"their":[19],"credibility":[20],"and":[21,61,64,173],"application":[22],"potential.":[23],"Existing":[24],"mitigation":[25],"methods":[26,82],"typically":[27],"rely":[28],"on":[29,151,181],"external":[30],"tools":[31],"or":[32],"the":[33,55,67,72,90,110,114,120,138,142,155,176,182],"comparison":[34],"of":[35,58,93,113,141,157,179],"multi-round":[36],"inference,":[37],"which":[38,51,105,136],"increase":[40],"inference":[41],"time.":[42],"In":[43],"this":[44,97],"paper,":[45],"we":[46,77,99,130],"propose":[47,100],"SElf-Evolving":[48],"Distillation":[49],"(SEED),":[50],"identifies":[52],"hallucinations":[53,165],"within":[54],"inner":[56],"knowledge":[57,69,116,140],"LVLMs,":[59],"isolates":[60],"purges":[62],"them,":[63],"then":[65],"distills":[66],"purified":[68,115,147],"back":[70],"into":[71],"model,":[73],"enabling":[74],"self-evolution.":[75],"Furthermore,":[76],"identified":[78],"that":[79,123],"traditional":[80],"distillation":[81,107],"are":[83],"prone":[84],"to":[85,108,190],"inducing":[86],"void":[87,127],"spaces":[88],"output":[91],"space":[92],"LVLMs.":[94],"To":[95],"address":[96],"issue,":[98],"a":[101,132],"Mode-Seeking":[102],"Evolving":[103],"approach,":[104],"performs":[106],"capture":[109],"dominant":[111],"modes":[112],"distribution,":[117],"thereby":[118],"avoiding":[119],"chaotic":[121],"results":[122],"could":[124],"emerge":[125],"from":[126,188],"spaces.":[128],"Moreover,":[129],"introduce":[131],"Hallucination":[133],"Elimination":[134],"Adapter,":[135],"corrects":[137],"dark":[139],"original":[143],"model":[144],"by":[145],"learning":[146],"knowledge.":[148],"Extensive":[149],"experiments":[150],"multiple":[152],"benchmarks":[153],"validate":[154],"superiority":[156],"our":[158],"SEED,":[159],"demonstrating":[160],"substantial":[161],"improvements":[162],"mitigating":[164],"for":[166],"representative":[167],"LVLM":[168],"models":[169],"LLaVA-1.5":[172,180],"InternVL2.":[174],"Remarkably,":[175],"F1":[177],"score":[178],"evaluation":[184],"metric":[185],"POPE-Random":[186],"improved":[187],"81.3":[189],"88.3.":[191]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-25T00:00:00"}