{"id":"https://openalex.org/W7147348305","doi":"https://doi.org/10.48550/arxiv.2603.29328","title":"Beyond Corner Patches: Semantics-Aware Backdoor Attack in Federated Learning","display_name":"Beyond Corner Patches: Semantics-Aware Backdoor Attack in Federated Learning","publication_year":2026,"publication_date":"2026-03-31","ids":{"openalex":"https://openalex.org/W7147348305","doi":"https://doi.org/10.48550/arxiv.2603.29328"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.29328","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.29328","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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.29328","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5132613984","display_name":"Kavindu Herath","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Herath, Kavindu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132656685","display_name":"Joshua Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Joshua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5047310442","display_name":"Saurabh Bagchi","orcid":"https://orcid.org/0000-0002-4239-5632"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bagchi, Saurabh","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":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":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.7595000267028809,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.7595000267028809,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.1534000039100647,"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":0.009399999864399433,"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/backdoor","display_name":"Backdoor","score":0.9954000115394592},{"id":"https://openalex.org/keywords/federated-learning","display_name":"Federated learning","score":0.5892000198364258},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.582099974155426},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.3885999917984009},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.33559998869895935},{"id":"https://openalex.org/keywords/synthetic-data","display_name":"Synthetic data","score":0.32409998774528503},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.29499998688697815}],"concepts":[{"id":"https://openalex.org/C2781045450","wikidata":"https://www.wikidata.org/wiki/Q254569","display_name":"Backdoor","level":2,"score":0.9954000115394592},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7351999878883362},{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.5892000198364258},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.582099974155426},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.4269999861717224},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41839998960494995},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.3885999917984009},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.33559998869895935},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.33070001006126404},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.32409998774528503},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.30640000104904175},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.29499998688697815},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.27480000257492065},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.27059999108314514},{"id":"https://openalex.org/C204806902","wikidata":"https://www.wikidata.org/wiki/Q2333581","display_name":"Semantic security","level":5,"score":0.2639000117778778},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C174333608","wikidata":"https://www.wikidata.org/wiki/Q19635","display_name":"Trojan","level":2,"score":0.25429999828338623}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.29328","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.29328","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.29328","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.29328","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":"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Backdoor":[0,60],"attacks":[1],"on":[2,101,181],"federated":[3,64,173],"learning":[4],"(FL)":[5],"are":[6,19],"most":[7],"often":[8],"evaluated":[9],"with":[10,84],"synthetic":[11,182],"corner":[12],"patches":[13],"or":[14],"out-of-distribution":[15],"(OOD)":[16],"patterns":[17],"that":[18,163,176],"unlikely":[20],"to":[21,33,90,95],"arise":[22],"in":[23,63,172],"practice.":[24],"In":[25],"this":[26],"paper,":[27],"we":[28],"revisit":[29],"the":[30,106,128],"backdoor":[31],"threat":[32,171],"standard":[34,129],"FL":[35,130],"(a":[36],"single":[37],"global":[38],"model)":[39],"under":[40],"a":[41,58,115,167],"more":[42],"realistic":[43],"setting":[44],"where":[45],"triggers":[46,70,148,184],"must":[47],"be":[48,186],"semantically":[49],"meaningful,":[50],"in-distribution,":[51],"and":[52,78,87,105,136,144,169,175],"visually":[53],"plausible.":[54],"We":[55,98],"propose":[56],"SABLE,":[57],"Semantics-Aware":[59],"for":[61],"LEarning":[62],"settings,":[65],"which":[66],"constructs":[67],"natural,":[68],"content-consistent":[69],"(e.g.,":[71],"semantic":[72],"attribute":[73],"changes":[74],"such":[75],"as":[76],"sunglasses)":[77],"optimizes":[79],"an":[80],"aggregation-aware":[81],"malicious":[82,121],"objective":[83],"feature":[85],"separation":[86],"parameter":[88],"regularization":[89],"keep":[91],"attacker":[92],"updates":[93],"close":[94],"benign":[96,157],"ones.":[97],"instantiate":[99],"SABLE":[100],"CelebA":[102],"hair-color":[103],"classification":[104],"German":[107],"Traffic":[108],"Sign":[109],"Recognition":[110],"Benchmark":[111],"(GTSRB),":[112],"poisoning":[113],"only":[114],"small,":[116],"interpretable":[117],"subset":[118],"of":[119],"each":[120],"client's":[122],"local":[123],"data":[124],"while":[125,155],"otherwise":[126],"following":[127],"protocol.":[131],"Across":[132],"heterogeneous":[133],"client":[134],"partitions":[135],"multiple":[137],"aggregation":[138],"rules":[139],"(FedAvg,":[140],"Trimmed":[141],"Mean,":[142],"MultiKrum,":[143],"FLAME),":[145],"our":[146],"semantics-driven":[147],"achieve":[149],"high":[150],"targeted":[151],"attack":[152],"success":[153],"rates":[154],"preserving":[156],"test":[158],"accuracy.":[159],"These":[160],"results":[161],"show":[162],"semantics-aligned":[164],"backdoors":[165],"remain":[166],"potent":[168],"practical":[170],"learning,":[174],"robustness":[177],"claims":[178],"based":[179],"solely":[180],"patch":[183],"can":[185],"overly":[187],"optimistic.":[188]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-02T00:00:00"}