Matches in SemOpenAlex for { <https://semopenalex.org/work/W3012909934> ?p ?o ?g. }
- W3012909934 abstract "Machine learning models, especially deep neural networks (DNNs), have been shown to be vulnerable against adversarial examples which are carefully crafted samples with a small magnitude of the perturbation. Such adversarial perturbations are usually restricted by bounding their $mathcal{L}_p$ norm such that they are imperceptible, and thus many current defenses can exploit this property to reduce their adversarial impact. In this paper, we instead introduce unrestricted perturbations that manipulate semantically meaningful image-based visual descriptors - color and texture - in order to generate effective and photorealistic adversarial examples. We show that these semantically aware perturbations are effective against JPEG compression, feature squeezing and adversarially trained model. We also show that the proposed methods can effectively be applied to both image classification and image captioning tasks on complex datasets such as ImageNet and MSCOCO. In addition, we conduct comprehensive user studies to show that our generated semantic adversarial examples are photorealistic to humans despite large magnitude perturbations when compared to other attacks." @default.
- W3012909934 created "2020-03-27" @default.
- W3012909934 creator A5009262110 @default.
- W3012909934 creator A5010851725 @default.
- W3012909934 creator A5028193823 @default.
- W3012909934 creator A5033616504 @default.
- W3012909934 creator A5062351534 @default.
- W3012909934 date "2019-04-12" @default.
- W3012909934 modified "2023-09-27" @default.
- W3012909934 title "Unrestricted Adversarial Examples via Semantic Manipulation" @default.
- W3012909934 cites W1686810756 @default.
- W3012909934 cites W1861492603 @default.
- W3012909934 cites W1905882502 @default.
- W3012909934 cites W1959608418 @default.
- W3012909934 cites W1999360130 @default.
- W3012909934 cites W2051434435 @default.
- W3012909934 cites W2064675550 @default.
- W3012909934 cites W2099471712 @default.
- W3012909934 cites W2108598243 @default.
- W3012909934 cites W2160815625 @default.
- W3012909934 cites W2163605009 @default.
- W3012909934 cites W2194775991 @default.
- W3012909934 cites W2326925005 @default.
- W3012909934 cites W2331128040 @default.
- W3012909934 cites W2408141691 @default.
- W3012909934 cites W2475287302 @default.
- W3012909934 cites W2584890299 @default.
- W3012909934 cites W2603777577 @default.
- W3012909934 cites W2612063021 @default.
- W3012909934 cites W2612372205 @default.
- W3012909934 cites W2783555701 @default.
- W3012909934 cites W2783784437 @default.
- W3012909934 cites W2798270931 @default.
- W3012909934 cites W2890591829 @default.
- W3012909934 cites W2890954269 @default.
- W3012909934 cites W2893749619 @default.
- W3012909934 cites W2937527502 @default.
- W3012909934 cites W2944600500 @default.
- W3012909934 cites W2952610664 @default.
- W3012909934 cites W2954182683 @default.
- W3012909934 cites W2962760235 @default.
- W3012909934 cites W2962772087 @default.
- W3012909934 cites W2963073614 @default.
- W3012909934 cites W2963207607 @default.
- W3012909934 cites W2963542245 @default.
- W3012909934 cites W2963594498 @default.
- W3012909934 cites W2963686907 @default.
- W3012909934 cites W2963857521 @default.
- W3012909934 cites W2964077693 @default.
- W3012909934 cites W2964121744 @default.
- W3012909934 cites W2964161785 @default.
- W3012909934 cites W2964193438 @default.
- W3012909934 cites W2969454448 @default.
- W3012909934 cites W3107942585 @default.
- W3012909934 cites W3024069363 @default.
- W3012909934 hasPublicationYear "2019" @default.
- W3012909934 type Work @default.
- W3012909934 sameAs 3012909934 @default.
- W3012909934 citedByCount "4" @default.
- W3012909934 countsByYear W30129099342021 @default.
- W3012909934 crossrefType "posted-content" @default.
- W3012909934 hasAuthorship W3012909934A5009262110 @default.
- W3012909934 hasAuthorship W3012909934A5010851725 @default.
- W3012909934 hasAuthorship W3012909934A5028193823 @default.
- W3012909934 hasAuthorship W3012909934A5033616504 @default.
- W3012909934 hasAuthorship W3012909934A5062351534 @default.
- W3012909934 hasConcept C111472728 @default.
- W3012909934 hasConcept C115961682 @default.
- W3012909934 hasConcept C119857082 @default.
- W3012909934 hasConcept C138885662 @default.
- W3012909934 hasConcept C154945302 @default.
- W3012909934 hasConcept C157657479 @default.
- W3012909934 hasConcept C165696696 @default.
- W3012909934 hasConcept C17744445 @default.
- W3012909934 hasConcept C189950617 @default.
- W3012909934 hasConcept C191795146 @default.
- W3012909934 hasConcept C199539241 @default.
- W3012909934 hasConcept C2984842247 @default.
- W3012909934 hasConcept C37736160 @default.
- W3012909934 hasConcept C38652104 @default.
- W3012909934 hasConcept C41008148 @default.
- W3012909934 hasConcept C50644808 @default.
- W3012909934 hasConcept C63584917 @default.
- W3012909934 hasConceptScore W3012909934C111472728 @default.
- W3012909934 hasConceptScore W3012909934C115961682 @default.
- W3012909934 hasConceptScore W3012909934C119857082 @default.
- W3012909934 hasConceptScore W3012909934C138885662 @default.
- W3012909934 hasConceptScore W3012909934C154945302 @default.
- W3012909934 hasConceptScore W3012909934C157657479 @default.
- W3012909934 hasConceptScore W3012909934C165696696 @default.
- W3012909934 hasConceptScore W3012909934C17744445 @default.
- W3012909934 hasConceptScore W3012909934C189950617 @default.
- W3012909934 hasConceptScore W3012909934C191795146 @default.
- W3012909934 hasConceptScore W3012909934C199539241 @default.
- W3012909934 hasConceptScore W3012909934C2984842247 @default.
- W3012909934 hasConceptScore W3012909934C37736160 @default.
- W3012909934 hasConceptScore W3012909934C38652104 @default.
- W3012909934 hasConceptScore W3012909934C41008148 @default.
- W3012909934 hasConceptScore W3012909934C50644808 @default.
- W3012909934 hasConceptScore W3012909934C63584917 @default.