Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385261657> ?p ?o ?g. }
Showing items 1 to 51 of
51
with 100 items per page.
- W4385261657 abstract "Unrestricted adversarial attacks present a serious threat to deep learning models and adversarial defense techniques. They pose severe security problems for deep learning applications because they can effectively bypass defense mechanisms. However, previous attack methods often utilize Generative Adversarial Networks (GANs), which are not theoretically provable and thus generate unrealistic examples by incorporating adversarial objectives, especially for large-scale datasets like ImageNet. In this paper, we propose a new method, called AdvDiff, to generate unrestricted adversarial examples with diffusion models. We design two novel adversarial guidance techniques to conduct adversarial sampling in the reverse generation process of diffusion models. These two techniques are effective and stable to generate high-quality, realistic adversarial examples by integrating gradients of the target classifier interpretably. Experimental results on MNIST and ImageNet datasets demonstrate that AdvDiff is effective to generate unrestricted adversarial examples, which outperforms GAN-based methods in terms of attack performance and generation quality." @default.
- W4385261657 created "2023-07-26" @default.
- W4385261657 creator A5024408463 @default.
- W4385261657 creator A5074934121 @default.
- W4385261657 creator A5076759808 @default.
- W4385261657 date "2023-07-23" @default.
- W4385261657 modified "2023-09-30" @default.
- W4385261657 title "AdvDiff: Generating Unrestricted Adversarial Examples using Diffusion Models" @default.
- W4385261657 doi "https://doi.org/10.48550/arxiv.2307.12499" @default.
- W4385261657 hasPublicationYear "2023" @default.
- W4385261657 type Work @default.
- W4385261657 citedByCount "0" @default.
- W4385261657 crossrefType "posted-content" @default.
- W4385261657 hasAuthorship W4385261657A5024408463 @default.
- W4385261657 hasAuthorship W4385261657A5074934121 @default.
- W4385261657 hasAuthorship W4385261657A5076759808 @default.
- W4385261657 hasBestOaLocation W43852616571 @default.
- W4385261657 hasConcept C108583219 @default.
- W4385261657 hasConcept C119857082 @default.
- W4385261657 hasConcept C154945302 @default.
- W4385261657 hasConcept C190502265 @default.
- W4385261657 hasConcept C2988773926 @default.
- W4385261657 hasConcept C37736160 @default.
- W4385261657 hasConcept C39890363 @default.
- W4385261657 hasConcept C41008148 @default.
- W4385261657 hasConcept C95623464 @default.
- W4385261657 hasConceptScore W4385261657C108583219 @default.
- W4385261657 hasConceptScore W4385261657C119857082 @default.
- W4385261657 hasConceptScore W4385261657C154945302 @default.
- W4385261657 hasConceptScore W4385261657C190502265 @default.
- W4385261657 hasConceptScore W4385261657C2988773926 @default.
- W4385261657 hasConceptScore W4385261657C37736160 @default.
- W4385261657 hasConceptScore W4385261657C39890363 @default.
- W4385261657 hasConceptScore W4385261657C41008148 @default.
- W4385261657 hasConceptScore W4385261657C95623464 @default.
- W4385261657 hasLocation W43852616571 @default.
- W4385261657 hasOpenAccess W4385261657 @default.
- W4385261657 hasPrimaryLocation W43852616571 @default.
- W4385261657 hasRelatedWork W2597787948 @default.
- W4385261657 hasRelatedWork W2759358869 @default.
- W4385261657 hasRelatedWork W2804545960 @default.
- W4385261657 hasRelatedWork W2972144487 @default.
- W4385261657 hasRelatedWork W2996316059 @default.
- W4385261657 hasRelatedWork W3000617323 @default.
- W4385261657 hasRelatedWork W3004372499 @default.
- W4385261657 hasRelatedWork W3156291593 @default.
- W4385261657 hasRelatedWork W3198184493 @default.
- W4385261657 hasRelatedWork W4220812973 @default.
- W4385261657 isParatext "false" @default.
- W4385261657 isRetracted "false" @default.
- W4385261657 workType "article" @default.