Matches in SemOpenAlex for { <https://semopenalex.org/work/W4380136140> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W4380136140 abstract "An important aspect of developing reliable deep learning systems is devising strategies that make these systems robust to adversarial attacks. There is a long line of work that focuses on developing defenses against these attacks, but recently, researchers have began to study ways to reverse engineer the attack process. This allows us to not only defend against several attack models, but also classify the threat model. However, there is still a lack of theoretical guarantees for the reverse engineering process. Current approaches that give any guarantees are based on the assumption that the data lies in a union of linear subspaces, which is not a valid assumption for more complex datasets. In this paper, we build on prior work and propose a novel framework for reverse engineering of deceptions which supposes that the clean data lies in the range of a GAN. To classify the signal and attack, we jointly solve a GAN inversion problem and a block-sparse recovery problem. For the first time in the literature, we provide deterministic linear convergence guarantees for this problem. We also empirically demonstrate the merits of the proposed approach on several nonlinear datasets as compared to state-of-the-art methods." @default.
- W4380136140 created "2023-06-10" @default.
- W4380136140 creator A5002186732 @default.
- W4380136140 creator A5011256828 @default.
- W4380136140 creator A5075524987 @default.
- W4380136140 date "2023-06-07" @default.
- W4380136140 modified "2023-10-09" @default.
- W4380136140 title "A Linearly Convergent GAN Inversion-based Algorithm for Reverse Engineering of Deceptions" @default.
- W4380136140 doi "https://doi.org/10.48550/arxiv.2306.04756" @default.
- W4380136140 hasPublicationYear "2023" @default.
- W4380136140 type Work @default.
- W4380136140 citedByCount "0" @default.
- W4380136140 crossrefType "posted-content" @default.
- W4380136140 hasAuthorship W4380136140A5002186732 @default.
- W4380136140 hasAuthorship W4380136140A5011256828 @default.
- W4380136140 hasAuthorship W4380136140A5075524987 @default.
- W4380136140 hasBestOaLocation W43801361401 @default.
- W4380136140 hasConcept C104317684 @default.
- W4380136140 hasConcept C109007969 @default.
- W4380136140 hasConcept C111919701 @default.
- W4380136140 hasConcept C11413529 @default.
- W4380136140 hasConcept C121332964 @default.
- W4380136140 hasConcept C12362212 @default.
- W4380136140 hasConcept C151730666 @default.
- W4380136140 hasConcept C154945302 @default.
- W4380136140 hasConcept C158622935 @default.
- W4380136140 hasConcept C162324750 @default.
- W4380136140 hasConcept C185592680 @default.
- W4380136140 hasConcept C1893757 @default.
- W4380136140 hasConcept C199360897 @default.
- W4380136140 hasConcept C207850805 @default.
- W4380136140 hasConcept C2524010 @default.
- W4380136140 hasConcept C2777210771 @default.
- W4380136140 hasConcept C2777303404 @default.
- W4380136140 hasConcept C33923547 @default.
- W4380136140 hasConcept C37736160 @default.
- W4380136140 hasConcept C41008148 @default.
- W4380136140 hasConcept C50522688 @default.
- W4380136140 hasConcept C55493867 @default.
- W4380136140 hasConcept C62520636 @default.
- W4380136140 hasConcept C63479239 @default.
- W4380136140 hasConcept C80444323 @default.
- W4380136140 hasConcept C86803240 @default.
- W4380136140 hasConcept C98045186 @default.
- W4380136140 hasConceptScore W4380136140C104317684 @default.
- W4380136140 hasConceptScore W4380136140C109007969 @default.
- W4380136140 hasConceptScore W4380136140C111919701 @default.
- W4380136140 hasConceptScore W4380136140C11413529 @default.
- W4380136140 hasConceptScore W4380136140C121332964 @default.
- W4380136140 hasConceptScore W4380136140C12362212 @default.
- W4380136140 hasConceptScore W4380136140C151730666 @default.
- W4380136140 hasConceptScore W4380136140C154945302 @default.
- W4380136140 hasConceptScore W4380136140C158622935 @default.
- W4380136140 hasConceptScore W4380136140C162324750 @default.
- W4380136140 hasConceptScore W4380136140C185592680 @default.
- W4380136140 hasConceptScore W4380136140C1893757 @default.
- W4380136140 hasConceptScore W4380136140C199360897 @default.
- W4380136140 hasConceptScore W4380136140C207850805 @default.
- W4380136140 hasConceptScore W4380136140C2524010 @default.
- W4380136140 hasConceptScore W4380136140C2777210771 @default.
- W4380136140 hasConceptScore W4380136140C2777303404 @default.
- W4380136140 hasConceptScore W4380136140C33923547 @default.
- W4380136140 hasConceptScore W4380136140C37736160 @default.
- W4380136140 hasConceptScore W4380136140C41008148 @default.
- W4380136140 hasConceptScore W4380136140C50522688 @default.
- W4380136140 hasConceptScore W4380136140C55493867 @default.
- W4380136140 hasConceptScore W4380136140C62520636 @default.
- W4380136140 hasConceptScore W4380136140C63479239 @default.
- W4380136140 hasConceptScore W4380136140C80444323 @default.
- W4380136140 hasConceptScore W4380136140C86803240 @default.
- W4380136140 hasConceptScore W4380136140C98045186 @default.
- W4380136140 hasLocation W43801361401 @default.
- W4380136140 hasOpenAccess W4380136140 @default.
- W4380136140 hasPrimaryLocation W43801361401 @default.
- W4380136140 hasRelatedWork W2069357446 @default.
- W4380136140 hasRelatedWork W2134680894 @default.
- W4380136140 hasRelatedWork W2155328232 @default.
- W4380136140 hasRelatedWork W2382095754 @default.
- W4380136140 hasRelatedWork W3094843325 @default.
- W4380136140 hasRelatedWork W4225161019 @default.
- W4380136140 hasRelatedWork W4307309570 @default.
- W4380136140 hasRelatedWork W4312326921 @default.
- W4380136140 hasRelatedWork W4315780078 @default.
- W4380136140 hasRelatedWork W3103180871 @default.
- W4380136140 isParatext "false" @default.
- W4380136140 isRetracted "false" @default.
- W4380136140 workType "article" @default.