Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890699875> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W2890699875 abstract "Recent studies have found that deep learning systems are vulnerable to adversarial examples; e.g., visually unrecognizable adversarial images can easily be crafted to result in misclassification. The robustness of neural networks has been studied extensively in the context of adversary detection, which compares a metric that exhibits strong discriminate power between natural and adversarial examples. In this paper, we propose to characterize the adversarial subspaces through the lens of mutual information (MI) approximated by conditional generation methods. We use MI as an information-theoretic metric to strengthen existing defenses and improve the performance of adversary detection. Experimental results on MagNet defense demonstrate that our proposed MI detector can strengthen its robustness against powerful adversarial attacks." @default.
- W2890699875 created "2018-09-27" @default.
- W2890699875 creator A5035387029 @default.
- W2890699875 creator A5044612822 @default.
- W2890699875 creator A5050344371 @default.
- W2890699875 creator A5085568918 @default.
- W2890699875 date "2018-11-01" @default.
- W2890699875 modified "2023-09-27" @default.
- W2890699875 title "ON THE UTILITY OF CONDITIONAL GENERATION BASED MUTUAL INFORMATION FOR CHARACTERIZING ADVERSARIAL SUBSPACES" @default.
- W2890699875 cites W2099471712 @default.
- W2890699875 cites W2243397390 @default.
- W2890699875 cites W2607219512 @default.
- W2890699875 cites W2618043096 @default.
- W2890699875 cites W2735471653 @default.
- W2890699875 cites W2766972025 @default.
- W2890699875 cites W2773446523 @default.
- W2890699875 cites W2794902890 @default.
- W2890699875 cites W2887603965 @default.
- W2890699875 cites W2963158386 @default.
- W2890699875 cites W2963207607 @default.
- W2890699875 cites W2963389226 @default.
- W2890699875 cites W2963564844 @default.
- W2890699875 cites W2963604280 @default.
- W2890699875 cites W2963612069 @default.
- W2890699875 cites W2963659778 @default.
- W2890699875 cites W2963857521 @default.
- W2890699875 cites W2964082701 @default.
- W2890699875 cites W2964153729 @default.
- W2890699875 cites W2964253222 @default.
- W2890699875 doi "https://doi.org/10.1109/globalsip.2018.8646527" @default.
- W2890699875 hasPublicationYear "2018" @default.
- W2890699875 type Work @default.
- W2890699875 sameAs 2890699875 @default.
- W2890699875 citedByCount "0" @default.
- W2890699875 crossrefType "proceedings-article" @default.
- W2890699875 hasAuthorship W2890699875A5035387029 @default.
- W2890699875 hasAuthorship W2890699875A5044612822 @default.
- W2890699875 hasAuthorship W2890699875A5050344371 @default.
- W2890699875 hasAuthorship W2890699875A5085568918 @default.
- W2890699875 hasBestOaLocation W28906998752 @default.
- W2890699875 hasConcept C104317684 @default.
- W2890699875 hasConcept C119857082 @default.
- W2890699875 hasConcept C12362212 @default.
- W2890699875 hasConcept C127413603 @default.
- W2890699875 hasConcept C152139883 @default.
- W2890699875 hasConcept C154945302 @default.
- W2890699875 hasConcept C176217482 @default.
- W2890699875 hasConcept C185592680 @default.
- W2890699875 hasConcept C21547014 @default.
- W2890699875 hasConcept C2524010 @default.
- W2890699875 hasConcept C33923547 @default.
- W2890699875 hasConcept C37736160 @default.
- W2890699875 hasConcept C38652104 @default.
- W2890699875 hasConcept C41008148 @default.
- W2890699875 hasConcept C41065033 @default.
- W2890699875 hasConcept C55493867 @default.
- W2890699875 hasConcept C63479239 @default.
- W2890699875 hasConcept C80444323 @default.
- W2890699875 hasConceptScore W2890699875C104317684 @default.
- W2890699875 hasConceptScore W2890699875C119857082 @default.
- W2890699875 hasConceptScore W2890699875C12362212 @default.
- W2890699875 hasConceptScore W2890699875C127413603 @default.
- W2890699875 hasConceptScore W2890699875C152139883 @default.
- W2890699875 hasConceptScore W2890699875C154945302 @default.
- W2890699875 hasConceptScore W2890699875C176217482 @default.
- W2890699875 hasConceptScore W2890699875C185592680 @default.
- W2890699875 hasConceptScore W2890699875C21547014 @default.
- W2890699875 hasConceptScore W2890699875C2524010 @default.
- W2890699875 hasConceptScore W2890699875C33923547 @default.
- W2890699875 hasConceptScore W2890699875C37736160 @default.
- W2890699875 hasConceptScore W2890699875C38652104 @default.
- W2890699875 hasConceptScore W2890699875C41008148 @default.
- W2890699875 hasConceptScore W2890699875C41065033 @default.
- W2890699875 hasConceptScore W2890699875C55493867 @default.
- W2890699875 hasConceptScore W2890699875C63479239 @default.
- W2890699875 hasConceptScore W2890699875C80444323 @default.
- W2890699875 hasLocation W28906998751 @default.
- W2890699875 hasLocation W28906998752 @default.
- W2890699875 hasOpenAccess W2890699875 @default.
- W2890699875 hasPrimaryLocation W28906998751 @default.
- W2890699875 hasRelatedWork W2610321374 @default.
- W2890699875 hasRelatedWork W2796814775 @default.
- W2890699875 hasRelatedWork W2805989439 @default.
- W2890699875 hasRelatedWork W2890699875 @default.
- W2890699875 hasRelatedWork W2951644823 @default.
- W2890699875 hasRelatedWork W2952919291 @default.
- W2890699875 hasRelatedWork W3035729345 @default.
- W2890699875 hasRelatedWork W4282972241 @default.
- W2890699875 hasRelatedWork W4287760660 @default.
- W2890699875 hasRelatedWork W9657784 @default.
- W2890699875 isParatext "false" @default.
- W2890699875 isRetracted "false" @default.
- W2890699875 magId "2890699875" @default.
- W2890699875 workType "article" @default.