Matches in SemOpenAlex for { <https://semopenalex.org/work/W2904920705> ?p ?o ?g. }
- W2904920705 abstract "In recent years, there have been significant advances in deep learning applied to problems in high-level vision tasks (e.g. image classification, object detection, semantic segmentation etc.) which has been met with a great deal of success. State-of-the-art methods that have shown impressive results on recognition tasks typically share a common structure involving stage-wise encoding of the image, followed by a generic classifier. However, these architectures have been shown to be vulnerable to the adversarial perturbations which may undermine the security of the systems supported by deep neural nets. In this work, initially we present rigorous evaluation of adversarial attacks on recent deep learning models for two different high-level tasks (image classification and semantic segmentation). Then we propose a model and dataset independent approach to generate adversarial perturbation and also the transferability of perturbation across different datasets and tasks. Moreover, we analyze the effect of different network architectures which will aid future efforts in understanding and defending against adversarial perturbations. We perform comprehensive experiments on several standard image classification and segmentation datasets to demonstrate the effectiveness of our proposed approach." @default.
- W2904920705 created "2018-12-22" @default.
- W2904920705 creator A5057462369 @default.
- W2904920705 creator A5067361805 @default.
- W2904920705 creator A5072076034 @default.
- W2904920705 creator A5075894318 @default.
- W2904920705 date "2018-05-01" @default.
- W2904920705 modified "2023-09-25" @default.
- W2904920705 title "On the Robustness of Deep Learning Models to Universal Adversarial Attack" @default.
- W2904920705 cites W1536680647 @default.
- W2904920705 cites W1686810756 @default.
- W2904920705 cites W1903029394 @default.
- W2904920705 cites W1932198206 @default.
- W2904920705 cites W2108598243 @default.
- W2904920705 cites W2194775991 @default.
- W2904920705 cites W2243397390 @default.
- W2904920705 cites W2460937040 @default.
- W2904920705 cites W2535873859 @default.
- W2904920705 cites W2543927648 @default.
- W2904920705 cites W2560023338 @default.
- W2904920705 cites W2604505099 @default.
- W2904920705 cites W2613718673 @default.
- W2904920705 cites W2734506812 @default.
- W2904920705 cites W2738229973 @default.
- W2904920705 cites W2753588254 @default.
- W2904920705 cites W2963118571 @default.
- W2904920705 cites W2963150697 @default.
- W2904920705 cites W2963207607 @default.
- W2904920705 cites W2964082701 @default.
- W2904920705 cites W2964097310 @default.
- W2904920705 cites W2964153729 @default.
- W2904920705 cites W2964288706 @default.
- W2904920705 cites W3118608800 @default.
- W2904920705 doi "https://doi.org/10.1109/crv.2018.00018" @default.
- W2904920705 hasPublicationYear "2018" @default.
- W2904920705 type Work @default.
- W2904920705 sameAs 2904920705 @default.
- W2904920705 citedByCount "3" @default.
- W2904920705 countsByYear W29049207052019 @default.
- W2904920705 countsByYear W29049207052020 @default.
- W2904920705 countsByYear W29049207052021 @default.
- W2904920705 crossrefType "proceedings-article" @default.
- W2904920705 hasAuthorship W2904920705A5057462369 @default.
- W2904920705 hasAuthorship W2904920705A5067361805 @default.
- W2904920705 hasAuthorship W2904920705A5072076034 @default.
- W2904920705 hasAuthorship W2904920705A5075894318 @default.
- W2904920705 hasConcept C104317684 @default.
- W2904920705 hasConcept C108583219 @default.
- W2904920705 hasConcept C115961682 @default.
- W2904920705 hasConcept C119857082 @default.
- W2904920705 hasConcept C124504099 @default.
- W2904920705 hasConcept C140331021 @default.
- W2904920705 hasConcept C153180895 @default.
- W2904920705 hasConcept C154945302 @default.
- W2904920705 hasConcept C185592680 @default.
- W2904920705 hasConcept C2776151529 @default.
- W2904920705 hasConcept C2984842247 @default.
- W2904920705 hasConcept C37736160 @default.
- W2904920705 hasConcept C41008148 @default.
- W2904920705 hasConcept C50644808 @default.
- W2904920705 hasConcept C55493867 @default.
- W2904920705 hasConcept C61272859 @default.
- W2904920705 hasConcept C63479239 @default.
- W2904920705 hasConcept C75294576 @default.
- W2904920705 hasConcept C89600930 @default.
- W2904920705 hasConcept C95623464 @default.
- W2904920705 hasConceptScore W2904920705C104317684 @default.
- W2904920705 hasConceptScore W2904920705C108583219 @default.
- W2904920705 hasConceptScore W2904920705C115961682 @default.
- W2904920705 hasConceptScore W2904920705C119857082 @default.
- W2904920705 hasConceptScore W2904920705C124504099 @default.
- W2904920705 hasConceptScore W2904920705C140331021 @default.
- W2904920705 hasConceptScore W2904920705C153180895 @default.
- W2904920705 hasConceptScore W2904920705C154945302 @default.
- W2904920705 hasConceptScore W2904920705C185592680 @default.
- W2904920705 hasConceptScore W2904920705C2776151529 @default.
- W2904920705 hasConceptScore W2904920705C2984842247 @default.
- W2904920705 hasConceptScore W2904920705C37736160 @default.
- W2904920705 hasConceptScore W2904920705C41008148 @default.
- W2904920705 hasConceptScore W2904920705C50644808 @default.
- W2904920705 hasConceptScore W2904920705C55493867 @default.
- W2904920705 hasConceptScore W2904920705C61272859 @default.
- W2904920705 hasConceptScore W2904920705C63479239 @default.
- W2904920705 hasConceptScore W2904920705C75294576 @default.
- W2904920705 hasConceptScore W2904920705C89600930 @default.
- W2904920705 hasConceptScore W2904920705C95623464 @default.
- W2904920705 hasLocation W29049207051 @default.
- W2904920705 hasOpenAccess W2904920705 @default.
- W2904920705 hasPrimaryLocation W29049207051 @default.
- W2904920705 hasRelatedWork W2593768305 @default.
- W2904920705 hasRelatedWork W2777623412 @default.
- W2904920705 hasRelatedWork W2923474628 @default.
- W2904920705 hasRelatedWork W2947704577 @default.
- W2904920705 hasRelatedWork W2948636084 @default.
- W2904920705 hasRelatedWork W2950172415 @default.
- W2904920705 hasRelatedWork W2950634055 @default.
- W2904920705 hasRelatedWork W2950774971 @default.
- W2904920705 hasRelatedWork W2963073217 @default.
- W2904920705 hasRelatedWork W2963107255 @default.
- W2904920705 hasRelatedWork W2963855547 @default.