{"id":"https://openalex.org/W7162639903","doi":"https://doi.org/10.1109/tip.2026.3695647","title":"Enhancing Zero-Shot Adversarial Robustness of Vision\u2013Language Models With Training-Free Adaptive Feature Movement","display_name":"Enhancing Zero-Shot Adversarial Robustness of Vision\u2013Language Models With Training-Free Adaptive Feature Movement","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7162639903","doi":"https://doi.org/10.1109/tip.2026.3695647","pmid":"https://pubmed.ncbi.nlm.nih.gov/42208028"},"language":"en","primary_location":{"id":"doi:10.1109/tip.2026.3695647","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tip.2026.3695647","pdf_url":null,"source":{"id":"https://openalex.org/S4210173141","display_name":"IEEE Transactions on Image Processing","issn_l":"1057-7149","issn":["1057-7149","1941-0042"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Image Processing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5080382446","display_name":"B. Y. Tong","orcid":"https://orcid.org/0000-0002-2831-8008"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Baoshun Tong","raw_affiliation_strings":["School of Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":"https://orcid.org/0009-0009-7979-9321","affiliations":[{"raw_affiliation_string":"School of Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137253409","display_name":"Hanjiang Lai","orcid":"https://orcid.org/0000-0001-8057-6744"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hanjiang Lai","raw_affiliation_strings":["School of Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":"https://orcid.org/0000-0001-8057-6744","affiliations":[{"raw_affiliation_string":"School of Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104243949","display_name":"Pan Y","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yan Pan","raw_affiliation_strings":["School of Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-0466-3763","affiliations":[{"raw_affiliation_string":"School of Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137248053","display_name":"Jian Yin","orcid":"https://orcid.org/0000-0002-1214-5384"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jian Yin","raw_affiliation_strings":["School of Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-1214-5384","affiliations":[{"raw_affiliation_string":"School of Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5104020899","display_name":"L Y Liang","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liang Lin","raw_affiliation_strings":["School of Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":"https://orcid.org/0000-0003-2248-3755","affiliations":[{"raw_affiliation_string":"School of Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157773358"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.66916791,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"35","issue":null,"first_page":"5657","last_page":"5671"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.3246000111103058,"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":0.3246000111103058,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3118000030517578,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.1598999947309494,"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/robustness","display_name":"Robustness (evolution)","score":0.7046999931335449},{"id":"https://openalex.org/keywords/adversarial-system","display_name":"Adversarial system","score":0.5593000054359436},{"id":"https://openalex.org/keywords/feature-extraction","display_name":"Feature extraction","score":0.41260001063346863},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.4047999978065491},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.4047999978065491},{"id":"https://openalex.org/keywords/image-processing","display_name":"Image processing","score":0.34940001368522644}],"concepts":[{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.7046999931335449},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6711000204086304},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6341000199317932},{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.5593000054359436},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.4742000102996826},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.41260001063346863},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4047999978065491},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.4047999978065491},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.34940001368522644},{"id":"https://openalex.org/C2780226923","wikidata":"https://www.wikidata.org/wiki/Q929848","display_name":"Movement (music)","level":2,"score":0.33079999685287476},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32989999651908875},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.30469998717308044},{"id":"https://openalex.org/C106516650","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm design","level":2,"score":0.29980000853538513},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.2782000005245209},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.2732999920845032},{"id":"https://openalex.org/C124504099","wikidata":"https://www.wikidata.org/wiki/Q56933","display_name":"Image segmentation","level":3,"score":0.26840001344680786}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tip.2026.3695647","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tip.2026.3695647","pdf_url":null,"source":{"id":"https://openalex.org/S4210173141","display_name":"IEEE Transactions on Image Processing","issn_l":"1057-7149","issn":["1057-7149","1941-0042"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Image Processing","raw_type":"journal-article"},{"id":"pmid:42208028","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/42208028","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on image processing : a publication of the IEEE Signal Processing Society","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities","score":0.41603076457977295}],"awards":[{"id":"https://openalex.org/G6248665153","display_name":null,"funder_award_id":"2024A1515012006","funder_id":"https://openalex.org/F4320337111","funder_display_name":"Basic and Applied Basic Research Foundation of Guangdong Province"}],"funders":[{"id":"https://openalex.org/F4320337111","display_name":"Basic and Applied Basic Research Foundation of Guangdong Province","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Pre-trained":[0],"Vision-Language":[1],"Models":[2],"(VLMs)":[3],"have":[4],"demonstrated":[5],"strong":[6],"zero-shot":[7,52,61,69,91,162,168,192],"generalization":[8,163],"capabilities.":[9],"Despite":[10],"their":[11,29],"effectiveness":[12],"on":[13,36,74,143,172,182],"various":[14],"downstream":[15],"tasks,":[16],"they":[17],"remain":[18],"vulnerable":[19],"to":[20,27,44,107,121,126,177],"adversarial":[21,34,70,97,119,150],"samples.":[22,124],"Existing":[23],"methods":[24],"fine-tune":[25],"VLMs":[26,73],"improve":[28,67],"robust":[30,193,198],"performance":[31],"by":[32,132,200],"performing":[33],"training":[35],"a":[37,50,59,109,166],"certain":[38],"dataset.":[39],"However,":[40],"this":[41,55],"can":[42,66,87,189],"lead":[43],"model":[45],"overfitting":[46,129],"and":[47,62,105,149,169],"is":[48],"not":[49],"true":[51],"scenario.":[53],"In":[54],"paper,":[56],"we":[57,79,94,135],"propose":[58,136],"truly":[60,167],"training-free":[63,170],"approach":[64],"that":[65,82,115,186],"the":[68,75,89,96,112,118,122,128,144,147,153,159,173,196],"robustness":[71],"of":[72,203],"evaluated":[76,174],"benchmarks.":[77],"Specifically,":[78],"first":[80],"discover":[81],"simply":[83],"adding":[84],"Gaussian":[85,101],"noise":[86,102],"enhance":[88],"VLM's":[90],"robustness.":[92],"Then,":[93],"treat":[95],"examples":[98,120],"with":[99],"added":[100],"as":[103],"anchors":[104,148],"strive":[106],"find":[108],"path":[110],"in":[111,152,165],"embedding":[113,154],"space":[114],"leads":[116],"from":[117],"cleaner":[123],"Furthermore,":[125],"avoid":[127],"issue":[130],"caused":[131],"fixed":[133],"hyperparameters,":[134],"an":[137,201],"adaptive":[138],"parameter":[139],"adjustment":[140],"method":[141,188],"based":[142],"distance":[145],"between":[146],"samples":[151],"space.":[155],"We":[156],"largely":[157],"preserve":[158],"original":[160],"VLMs'":[161],"abilities":[164],"manner":[171],"benchmarks":[175],"compared":[176],"previous":[178],"methods.":[179],"Extensive":[180],"experiments":[181],"16":[183],"datasets":[184],"demonstrate":[185],"our":[187],"achieve":[190],"stronger":[191],"performance,":[194],"improving":[195],"top-1":[197],"accuracy":[199],"average":[202],"10.83%.":[204]},"counts_by_year":[],"updated_date":"2026-06-03T06:16:58.514037","created_date":"2026-05-29T00:00:00"}