Matches in SemOpenAlex for { <https://semopenalex.org/work/W2892020710> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W2892020710 abstract "In the last decade, deep learning algorithms have become very popular thanks to the achieved performance in many machine learning and computer vision tasks. However, most of the deep learning architectures are vulnerable to so called adversarial examples. This questions the security of deep neural networks (DNN) for many security- and trust-sensitive domains. The majority of the proposed existing adversarial attacks are based on the differentiability of the DNN cost function.Defence strategies are mostly based on machine learning and signal processing principles that either try to detect-reject or filter out the adversarial perturbations and completely neglect the classical cryptographic component in the defence. In this work, we propose a new defence mechanism based on the second Kerckhoffs's cryptographic principle which states that the defence and classification algorithm are supposed to be known, but not the key. To be compliant with the assumption that the attacker does not have access to the secret key, we will primarily focus on a gray-box scenario and do not address a white-box one. More particularly, we assume that the attacker does not have direct access to the secret block, but (a) he completely knows the system architecture, (b) he has access to the data used for training and testing and (c) he can observe the output of the classifier for each given input. We show empirically that our system is efficient against most famous state-of-the-art attacks in black-box and gray-box scenarios." @default.
- W2892020710 created "2018-09-27" @default.
- W2892020710 creator A5008678783 @default.
- W2892020710 creator A5026953476 @default.
- W2892020710 creator A5091506990 @default.
- W2892020710 date "2018-09-05" @default.
- W2892020710 modified "2023-10-05" @default.
- W2892020710 title "Bridging machine learning and cryptography in defence against adversarial attacks" @default.
- W2892020710 cites W1883420340 @default.
- W2892020710 cites W2180612164 @default.
- W2892020710 cites W2230740169 @default.
- W2892020710 cites W2243397390 @default.
- W2892020710 cites W2460937040 @default.
- W2892020710 cites W2561975083 @default.
- W2892020710 cites W2619479788 @default.
- W2892020710 cites W2625220439 @default.
- W2892020710 cites W2750384547 @default.
- W2892020710 cites W2754537581 @default.
- W2892020710 cites W2760600531 @default.
- W2892020710 cites W2777449390 @default.
- W2892020710 cites W2787708942 @default.
- W2892020710 cites W2902071009 @default.
- W2892020710 cites W2949311987 @default.
- W2892020710 cites W2950468330 @default.
- W2892020710 cites W2950906520 @default.
- W2892020710 cites W2963207607 @default.
- W2892020710 cites W2963695663 @default.
- W2892020710 cites W2963744840 @default.
- W2892020710 cites W2963857521 @default.
- W2892020710 cites W2964082701 @default.
- W2892020710 cites W2964153729 @default.
- W2892020710 hasPublicationYear "2018" @default.
- W2892020710 type Work @default.
- W2892020710 sameAs 2892020710 @default.
- W2892020710 citedByCount "0" @default.
- W2892020710 crossrefType "posted-content" @default.
- W2892020710 hasAuthorship W2892020710A5008678783 @default.
- W2892020710 hasAuthorship W2892020710A5026953476 @default.
- W2892020710 hasAuthorship W2892020710A5091506990 @default.
- W2892020710 hasConcept C108583219 @default.
- W2892020710 hasConcept C119857082 @default.
- W2892020710 hasConcept C147343967 @default.
- W2892020710 hasConcept C154945302 @default.
- W2892020710 hasConcept C174348530 @default.
- W2892020710 hasConcept C178489894 @default.
- W2892020710 hasConcept C180932941 @default.
- W2892020710 hasConcept C2778403875 @default.
- W2892020710 hasConcept C2984842247 @default.
- W2892020710 hasConcept C37736160 @default.
- W2892020710 hasConcept C38652104 @default.
- W2892020710 hasConcept C41008148 @default.
- W2892020710 hasConcept C80444323 @default.
- W2892020710 hasConcept C95623464 @default.
- W2892020710 hasConceptScore W2892020710C108583219 @default.
- W2892020710 hasConceptScore W2892020710C119857082 @default.
- W2892020710 hasConceptScore W2892020710C147343967 @default.
- W2892020710 hasConceptScore W2892020710C154945302 @default.
- W2892020710 hasConceptScore W2892020710C174348530 @default.
- W2892020710 hasConceptScore W2892020710C178489894 @default.
- W2892020710 hasConceptScore W2892020710C180932941 @default.
- W2892020710 hasConceptScore W2892020710C2778403875 @default.
- W2892020710 hasConceptScore W2892020710C2984842247 @default.
- W2892020710 hasConceptScore W2892020710C37736160 @default.
- W2892020710 hasConceptScore W2892020710C38652104 @default.
- W2892020710 hasConceptScore W2892020710C41008148 @default.
- W2892020710 hasConceptScore W2892020710C80444323 @default.
- W2892020710 hasConceptScore W2892020710C95623464 @default.
- W2892020710 hasLocation W28920207101 @default.
- W2892020710 hasOpenAccess W2892020710 @default.
- W2892020710 hasPrimaryLocation W28920207101 @default.
- W2892020710 hasRelatedWork W2640329709 @default.
- W2892020710 hasRelatedWork W2773022113 @default.
- W2892020710 hasRelatedWork W2886165587 @default.
- W2892020710 hasRelatedWork W2886757827 @default.
- W2892020710 hasRelatedWork W2949311987 @default.
- W2892020710 hasRelatedWork W2962878175 @default.
- W2892020710 hasRelatedWork W2963448658 @default.
- W2892020710 hasRelatedWork W2964277985 @default.
- W2892020710 hasRelatedWork W2964286909 @default.
- W2892020710 hasRelatedWork W2972619028 @default.
- W2892020710 hasRelatedWork W2979438533 @default.
- W2892020710 hasRelatedWork W3047856150 @default.
- W2892020710 hasRelatedWork W3091857398 @default.
- W2892020710 hasRelatedWork W3093482919 @default.
- W2892020710 hasRelatedWork W3100485466 @default.
- W2892020710 hasRelatedWork W3102139703 @default.
- W2892020710 hasRelatedWork W3105389675 @default.
- W2892020710 hasRelatedWork W3108822760 @default.
- W2892020710 hasRelatedWork W3112205328 @default.
- W2892020710 hasRelatedWork W3207985630 @default.
- W2892020710 isParatext "false" @default.
- W2892020710 isRetracted "false" @default.
- W2892020710 magId "2892020710" @default.
- W2892020710 workType "article" @default.