{"id":"https://openalex.org/W4416034174","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.1025","title":"MaskCD: Mitigating LVLM Hallucinations by Image Head Masked Contrastive Decoding","display_name":"MaskCD: Mitigating LVLM Hallucinations by Image Head Masked Contrastive Decoding","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416034174","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.1025"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.1025","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1025","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1025.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":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.1025.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103223538","display_name":"Jingyuan Deng","orcid":"https://orcid.org/0009-0000-1225-8947"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jingyuan Deng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5020953714","display_name":"Yujiu Yang","orcid":"https://orcid.org/0000-0002-6427-1024"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yujiu Yang","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.28839442,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"18854","last_page":"18866"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13397","display_name":"Hallucinations in medical conditions","score":0.39750000834465027,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},"topics":[{"id":"https://openalex.org/T13397","display_name":"Hallucinations in medical conditions","score":0.39750000834465027,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.051100000739097595,"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/T12553","display_name":"Psychedelics and Drug Studies","score":0.03970000147819519,"subfield":{"id":"https://openalex.org/subfields/3203","display_name":"Clinical Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/head","display_name":"Head (geology)","score":0.5073999762535095},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.5034000277519226},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.3375999927520752},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.30570000410079956},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.28999999165534973},{"id":"https://openalex.org/keywords/image-processing","display_name":"Image processing","score":0.2840999960899353}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.569100022315979},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5339000225067139},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.5180000066757202},{"id":"https://openalex.org/C2780312720","wikidata":"https://www.wikidata.org/wiki/Q5689100","display_name":"Head (geology)","level":2,"score":0.5073999762535095},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5034000277519226},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3391999900341034},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.3375999927520752},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.30570000410079956},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2906000018119812},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.28999999165534973},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.2840999960899353},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.273499995470047},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.2637999951839447},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.259799987077713},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.1025","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1025","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1025.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.1025","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1025","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1025.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":[],"funders":[{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416034174.pdf","grobid_xml":"https://content.openalex.org/works/W4416034174.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"vision-language":[1],"models":[2],"(LVLMs)":[3],"have":[4,26,48],"shown":[5],"remarkable":[6],"performance":[7],"in":[8,67,96],"visuallanguage":[9],"understanding":[10],"for":[11,32,104],"downstream":[12],"multimodal":[13],"tasks.While":[14],"their":[15,41],"capabilities":[16,137],"are":[17,76],"improving,":[18],"problems":[19],"emerge":[20],"simultaneously.Among":[21],"those":[22],"problems,":[23],"the":[24,33,93,129,135],"hallucinations":[25,132],"attracted":[27],"much":[28],"attention,":[29],"which":[30],"stands":[31],"phenomenon":[34,130],"where":[35],"LVLMs":[36],"generate":[37],"contradictory":[38],"content":[39],"to":[40,51,100],"input":[42],"visual":[43],"and":[44,60,72,111,121,133],"text":[45],"contents.Many":[46],"approaches":[47],"been":[49],"proposed":[50],"deal":[52],"with":[53],"this":[54,81],"issue,":[55],"such":[56,116],"as":[57,117],"contrastive":[58,63,70,102,105],"decoding":[59,64],"attention":[61,73],"manipulation.However,":[62],"methods":[65,75],"struggle":[66],"constructing":[68],"appropriate":[69],"samples,":[71],"manipulation":[74],"highly":[77],"sensitive,":[78],"lacking":[79],"stability.In":[80],"work,":[82],"we":[83],"propose":[84],"image":[85],"head":[86],"Masked":[87],"Contrastive":[88],"Decoding":[89],"(MaskCD).Our":[90],"approach":[91],"utilizes":[92],"\"image":[94],"heads\"":[95],"LVLMs,":[97],"masking":[98],"them":[99],"construct":[101],"samples":[103],"decoding.We":[106],"evaluated":[107],"MaskCD":[108,126],"on":[109],"LLaVA-1.5-7b":[110],"Qwen-VL-7b,":[112],"using":[113],"various":[114],"benchmarks":[115],"CHAIR,":[118],"POPE,":[119],"AMBER":[120],"MME.The":[122],"results":[123],"demonstrate":[124],"that":[125],"effectively":[127],"alleviates":[128],"of":[131,138],"retains":[134],"general":[136],"LVLMs.":[139]},"counts_by_year":[],"updated_date":"2026-03-12T06:13:28.667946","created_date":"2025-11-08T00:00:00"}