Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386025614> ?p ?o ?g. }
- W4386025614 endingPage "89382" @default.
- W4386025614 startingPage "89369" @default.
- W4386025614 abstract "Deep face recognition models are vulnerable to adversarial samples generated by adversarial attack methods. However, current attack methods do not adequately represent the security problems of the deep FR models, because they either produce adversarial samples which are unnatural and easily perceived by human or have poor attack capabilities with low attack success rates on the black-box victim FR model. To achieve a good trade-off between the imperceptibility and attack capability, we propose Adv-Eye, a novel method for constructing adversarial facial images by adding natural eyeshadow to the orbital region. Adv-Eye consists of Makeup Generation Module, Makeup Blending Module, and Attack Module. Makeup Generation Module develops pre-makeup strategy to help adversarial generative networks (GANs) to accurately generate eyeshadow on the orbital image. Makeup Blending Module develops a multi-view image visual similarity evaluation method to improve the imperceptibility of the generated eyeshadow. In Attack Module, an ensemble attack strategy based on fine-grained meta-learning and input decay, is applied to improve attack capability under query-free black-box setting. From the experimental results under LADN dataset and MT dataset, compared with existing techniques, the adversarial samples generated by Adv-Eye not only significantly improve the visual quality, but also get 1.63% and 1.05% average success rates improving on the local black-box FR model, 5.33 and 5.22 average confidence point improving on the online commercial FR plateform respectively. The above results demonstrate that pre-makeup strategy and multi-view image visual similarity evaluation method effectively improve the imperceptibility of generated adversarial perturbations, and Attack Module effectively improves attack success rate while maintaining high image quality." @default.
- W4386025614 created "2023-08-22" @default.
- W4386025614 creator A5020796294 @default.
- W4386025614 creator A5023919878 @default.
- W4386025614 creator A5032483261 @default.
- W4386025614 creator A5037531369 @default.
- W4386025614 creator A5056134541 @default.
- W4386025614 creator A5057334944 @default.
- W4386025614 creator A5048210685 @default.
- W4386025614 date "2023-01-01" @default.
- W4386025614 modified "2023-09-26" @default.
- W4386025614 title "Adv-Eye: A Transfer-based Natural Eye Shadow Attack on Face Recognition" @default.
- W4386025614 cites W1964859077 @default.
- W4386025614 cites W2133665775 @default.
- W4386025614 cites W2194775991 @default.
- W4386025614 cites W2535873859 @default.
- W4386025614 cites W2603766943 @default.
- W4386025614 cites W2752782242 @default.
- W4386025614 cites W2774644650 @default.
- W4386025614 cites W2795727551 @default.
- W4386025614 cites W2798600195 @default.
- W4386025614 cites W2896240508 @default.
- W4386025614 cites W2932026309 @default.
- W4386025614 cites W2962770929 @default.
- W4386025614 cites W2962785568 @default.
- W4386025614 cites W2962793481 @default.
- W4386025614 cites W2962847335 @default.
- W4386025614 cites W2963073614 @default.
- W4386025614 cites W2963857521 @default.
- W4386025614 cites W2969542116 @default.
- W4386025614 cites W2969664989 @default.
- W4386025614 cites W2969985801 @default.
- W4386025614 cites W2970255644 @default.
- W4386025614 cites W2972986629 @default.
- W4386025614 cites W2989149111 @default.
- W4386025614 cites W2997095758 @default.
- W4386025614 cites W3005588871 @default.
- W4386025614 cites W3009134234 @default.
- W4386025614 cites W3035693354 @default.
- W4386025614 cites W3083540668 @default.
- W4386025614 cites W3097573595 @default.
- W4386025614 cites W3099206234 @default.
- W4386025614 cites W3119391778 @default.
- W4386025614 cites W3123981332 @default.
- W4386025614 cites W3127807678 @default.
- W4386025614 cites W3171288285 @default.
- W4386025614 cites W3179647175 @default.
- W4386025614 cites W3189402954 @default.
- W4386025614 cites W4207075225 @default.
- W4386025614 cites W4214502238 @default.
- W4386025614 cites W4224923640 @default.
- W4386025614 cites W4293157178 @default.
- W4386025614 cites W4312109322 @default.
- W4386025614 cites W4312937923 @default.
- W4386025614 cites W4319300575 @default.
- W4386025614 cites W4385446764 @default.
- W4386025614 doi "https://doi.org/10.1109/access.2023.3307132" @default.
- W4386025614 hasPublicationYear "2023" @default.
- W4386025614 type Work @default.
- W4386025614 citedByCount "0" @default.
- W4386025614 crossrefType "journal-article" @default.
- W4386025614 hasAuthorship W4386025614A5020796294 @default.
- W4386025614 hasAuthorship W4386025614A5023919878 @default.
- W4386025614 hasAuthorship W4386025614A5032483261 @default.
- W4386025614 hasAuthorship W4386025614A5037531369 @default.
- W4386025614 hasAuthorship W4386025614A5048210685 @default.
- W4386025614 hasAuthorship W4386025614A5056134541 @default.
- W4386025614 hasAuthorship W4386025614A5057334944 @default.
- W4386025614 hasBestOaLocation W43860256141 @default.
- W4386025614 hasConcept C103278499 @default.
- W4386025614 hasConcept C108583219 @default.
- W4386025614 hasConcept C115961682 @default.
- W4386025614 hasConcept C117797892 @default.
- W4386025614 hasConcept C119857082 @default.
- W4386025614 hasConcept C144024400 @default.
- W4386025614 hasConcept C153180895 @default.
- W4386025614 hasConcept C154945302 @default.
- W4386025614 hasConcept C15744967 @default.
- W4386025614 hasConcept C2524010 @default.
- W4386025614 hasConcept C2779304628 @default.
- W4386025614 hasConcept C28719098 @default.
- W4386025614 hasConcept C31510193 @default.
- W4386025614 hasConcept C31972630 @default.
- W4386025614 hasConcept C33923547 @default.
- W4386025614 hasConcept C36289849 @default.
- W4386025614 hasConcept C37736160 @default.
- W4386025614 hasConcept C41008148 @default.
- W4386025614 hasConcept C542102704 @default.
- W4386025614 hasConceptScore W4386025614C103278499 @default.
- W4386025614 hasConceptScore W4386025614C108583219 @default.
- W4386025614 hasConceptScore W4386025614C115961682 @default.
- W4386025614 hasConceptScore W4386025614C117797892 @default.
- W4386025614 hasConceptScore W4386025614C119857082 @default.
- W4386025614 hasConceptScore W4386025614C144024400 @default.
- W4386025614 hasConceptScore W4386025614C153180895 @default.
- W4386025614 hasConceptScore W4386025614C154945302 @default.
- W4386025614 hasConceptScore W4386025614C15744967 @default.
- W4386025614 hasConceptScore W4386025614C2524010 @default.