Matches in SemOpenAlex for { <https://semopenalex.org/work/W3163269931> ?p ?o ?g. }
- W3163269931 abstract "Due to its powerful capability of representation learning and high-efficiency computation, deep hashing has made significant progress in large-scale image retrieval. However, deep hashing networks are vulnerable to adversarial examples, which is a practical secure problem but seldom studied in hashing-based retrieval field. In this paper, we propose a novel prototype-supervised adversarial network (ProS-GAN), which formulates a flexible generative architecture for efficient and effective targeted hashing attack. To the best of our knowledge, this is the first generation-based method to attack deep hashing networks. Generally, our proposed framework consists of three parts, i.e., a PrototypeNet, a generator, and a discriminator. Specifically, the designed PrototypeNet embeds the target label into the semantic representation and learns the prototype code as the category-level representative of the target label. Moreover, the semantic representation and the original image are jointly fed into the generator for a flexible targeted attack. Particularly, the prototype code is adopted to supervise the generator to construct the targeted adversarial example by minimizing the Hamming distance between the hash code of the adversarial example and the prototype code. Furthermore, the generator is against the discriminator to simultaneously encourage the adversarial examples visually realistic and the semantic representation informative. Extensive experiments verify that the proposed framework can efficiently produce adversarial examples with better targeted attack performance and transferability over state-of-the-art targeted attack methods of deep hashing. The related codes could be available at https://github.com/xunguangwang/ProS-GAN ." @default.
- W3163269931 created "2021-05-24" @default.
- W3163269931 creator A5026836722 @default.
- W3163269931 creator A5030843117 @default.
- W3163269931 creator A5068027800 @default.
- W3163269931 creator A5074492050 @default.
- W3163269931 creator A5088888083 @default.
- W3163269931 date "2021-05-16" @default.
- W3163269931 modified "2023-10-16" @default.
- W3163269931 title "Prototype-supervised Adversarial Network for Targeted Attack of Deep Hashing" @default.
- W3163269931 cites W1861492603 @default.
- W3163269931 cites W1901129140 @default.
- W3163269931 cites W1939575207 @default.
- W3163269931 cites W2007972815 @default.
- W3163269931 cites W2099471712 @default.
- W3163269931 cites W2130660124 @default.
- W3163269931 cites W2145065594 @default.
- W3163269931 cites W2155803963 @default.
- W3163269931 cites W2163605009 @default.
- W3163269931 cites W2194775991 @default.
- W3163269931 cites W2266728343 @default.
- W3163269931 cites W2293824885 @default.
- W3163269931 cites W2411707397 @default.
- W3163269931 cites W2464915613 @default.
- W3163269931 cites W2508837377 @default.
- W3163269931 cites W2603766943 @default.
- W3163269931 cites W2604147826 @default.
- W3163269931 cites W2781821509 @default.
- W3163269931 cites W2798834175 @default.
- W3163269931 cites W2799245683 @default.
- W3163269931 cites W2809299793 @default.
- W3163269931 cites W2883233582 @default.
- W3163269931 cites W2904169782 @default.
- W3163269931 cites W2962786441 @default.
- W3163269931 cites W2962793481 @default.
- W3163269931 cites W2962835968 @default.
- W3163269931 cites W2963187862 @default.
- W3163269931 cites W2963207607 @default.
- W3163269931 cites W2963389226 @default.
- W3163269931 cites W2963398644 @default.
- W3163269931 cites W2963448658 @default.
- W3163269931 cites W2963560523 @default.
- W3163269931 cites W2963855547 @default.
- W3163269931 cites W2963857521 @default.
- W3163269931 cites W2964121744 @default.
- W3163269931 cites W2964153729 @default.
- W3163269931 cites W2964253222 @default.
- W3163269931 cites W2970652125 @default.
- W3163269931 cites W2984942115 @default.
- W3163269931 cites W2997223206 @default.
- W3163269931 cites W3006896454 @default.
- W3163269931 cites W3009756253 @default.
- W3163269931 cites W3027918202 @default.
- W3163269931 cites W3080270960 @default.
- W3163269931 cites W3086725547 @default.
- W3163269931 cites W3108980562 @default.
- W3163269931 doi "https://doi.org/10.48550/arxiv.2105.07553" @default.
- W3163269931 hasPublicationYear "2021" @default.
- W3163269931 type Work @default.
- W3163269931 sameAs 3163269931 @default.
- W3163269931 citedByCount "0" @default.
- W3163269931 crossrefType "posted-content" @default.
- W3163269931 hasAuthorship W3163269931A5026836722 @default.
- W3163269931 hasAuthorship W3163269931A5030843117 @default.
- W3163269931 hasAuthorship W3163269931A5068027800 @default.
- W3163269931 hasAuthorship W3163269931A5074492050 @default.
- W3163269931 hasAuthorship W3163269931A5088888083 @default.
- W3163269931 hasBestOaLocation W31632699311 @default.
- W3163269931 hasConcept C108583219 @default.
- W3163269931 hasConcept C113775141 @default.
- W3163269931 hasConcept C119857082 @default.
- W3163269931 hasConcept C121332964 @default.
- W3163269931 hasConcept C154945302 @default.
- W3163269931 hasConcept C163258240 @default.
- W3163269931 hasConcept C177264268 @default.
- W3163269931 hasConcept C17744445 @default.
- W3163269931 hasConcept C199360897 @default.
- W3163269931 hasConcept C199539241 @default.
- W3163269931 hasConcept C2776359362 @default.
- W3163269931 hasConcept C2776760102 @default.
- W3163269931 hasConcept C2779803651 @default.
- W3163269931 hasConcept C2780992000 @default.
- W3163269931 hasConcept C37736160 @default.
- W3163269931 hasConcept C38652104 @default.
- W3163269931 hasConcept C41008148 @default.
- W3163269931 hasConcept C62520636 @default.
- W3163269931 hasConcept C76155785 @default.
- W3163269931 hasConcept C80444323 @default.
- W3163269931 hasConcept C94625758 @default.
- W3163269931 hasConcept C94915269 @default.
- W3163269931 hasConcept C99138194 @default.
- W3163269931 hasConceptScore W3163269931C108583219 @default.
- W3163269931 hasConceptScore W3163269931C113775141 @default.
- W3163269931 hasConceptScore W3163269931C119857082 @default.
- W3163269931 hasConceptScore W3163269931C121332964 @default.
- W3163269931 hasConceptScore W3163269931C154945302 @default.
- W3163269931 hasConceptScore W3163269931C163258240 @default.
- W3163269931 hasConceptScore W3163269931C177264268 @default.
- W3163269931 hasConceptScore W3163269931C17744445 @default.
- W3163269931 hasConceptScore W3163269931C199360897 @default.