Matches in SemOpenAlex for { <https://semopenalex.org/work/W3204996710> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W3204996710 abstract "We propose a novel method for creating adversarial examples. Instead of perturbing pixels, we use an encoder-decoder representation of the input image and perturb intermediate layers in the decoder. This changes the high-level features provided by the generative model. Therefore, our perturbation possesses semantic meaning, such as a longer beak or green tints. We formulate this task as an optimization problem by minimizing the Wasserstein distance between the adversarial and initial images under a misclassification constraint. We employ the projected gradient method with a simple inexact projection. Due to the projection, all iterations are feasible, and our method always generates adversarial images. We perform numerical experiments on the MNIST and ImageNet datasets in both targeted and untargeted settings. We demonstrate that our adversarial images are much less vulnerable to steganographic defence techniques than pixel-based attacks. Moreover, we show that our method modifies key features such as edges and that defence techniques based on adversarial training are vulnerable to our attacks." @default.
- W3204996710 created "2021-10-25" @default.
- W3204996710 creator A5030525746 @default.
- W3204996710 creator A5071184886 @default.
- W3204996710 date "2022-01-01" @default.
- W3204996710 modified "2023-09-25" @default.
- W3204996710 title "Adversarial Examples by Perturbing High-level Features in Intermediate Decoder Layers" @default.
- W3204996710 cites W2108598243 @default.
- W3204996710 cites W2158131535 @default.
- W3204996710 cites W2180612164 @default.
- W3204996710 cites W2411541852 @default.
- W3204996710 cites W2412320034 @default.
- W3204996710 cites W2618043096 @default.
- W3204996710 cites W2897105053 @default.
- W3204996710 cites W2915401840 @default.
- W3204996710 cites W2955425717 @default.
- W3204996710 cites W2962835968 @default.
- W3204996710 cites W2963207607 @default.
- W3204996710 cites W2963389226 @default.
- W3204996710 cites W2963448658 @default.
- W3204996710 cites W2963857521 @default.
- W3204996710 cites W2964153729 @default.
- W3204996710 cites W2964197269 @default.
- W3204996710 cites W2964238361 @default.
- W3204996710 cites W2964253222 @default.
- W3204996710 cites W2970241862 @default.
- W3204996710 cites W3015625436 @default.
- W3204996710 cites W3029645440 @default.
- W3204996710 cites W3088909400 @default.
- W3204996710 cites W3142067920 @default.
- W3204996710 doi "https://doi.org/10.5220/0010892800003116" @default.
- W3204996710 hasPublicationYear "2022" @default.
- W3204996710 type Work @default.
- W3204996710 sameAs 3204996710 @default.
- W3204996710 citedByCount "0" @default.
- W3204996710 crossrefType "proceedings-article" @default.
- W3204996710 hasAuthorship W3204996710A5030525746 @default.
- W3204996710 hasAuthorship W3204996710A5071184886 @default.
- W3204996710 hasBestOaLocation W32049967101 @default.
- W3204996710 hasConcept C11413529 @default.
- W3204996710 hasConcept C154945302 @default.
- W3204996710 hasConcept C37736160 @default.
- W3204996710 hasConcept C41008148 @default.
- W3204996710 hasConceptScore W3204996710C11413529 @default.
- W3204996710 hasConceptScore W3204996710C154945302 @default.
- W3204996710 hasConceptScore W3204996710C37736160 @default.
- W3204996710 hasConceptScore W3204996710C41008148 @default.
- W3204996710 hasLocation W32049967101 @default.
- W3204996710 hasLocation W32049967102 @default.
- W3204996710 hasOpenAccess W3204996710 @default.
- W3204996710 hasPrimaryLocation W32049967101 @default.
- W3204996710 hasRelatedWork W1858327386 @default.
- W3204996710 hasRelatedWork W2386767533 @default.
- W3204996710 hasRelatedWork W2901368259 @default.
- W3204996710 hasRelatedWork W2903917280 @default.
- W3204996710 hasRelatedWork W3100356292 @default.
- W3204996710 hasRelatedWork W3156291593 @default.
- W3204996710 hasRelatedWork W3198184493 @default.
- W3204996710 hasRelatedWork W4205705013 @default.
- W3204996710 hasRelatedWork W4295850094 @default.
- W3204996710 hasRelatedWork W4304891817 @default.
- W3204996710 isParatext "false" @default.
- W3204996710 isRetracted "false" @default.
- W3204996710 magId "3204996710" @default.
- W3204996710 workType "article" @default.