Matches in SemOpenAlex for { <https://semopenalex.org/work/W4362609169> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4362609169 endingPage "4599" @default.
- W4362609169 startingPage "4599" @default.
- W4362609169 abstract "The successful outcomes of deep learning (DL) algorithms in diverse fields have prompted researchers to consider backdoor attacks on DL models to defend them in practical applications. Adversarial examples could deceive a safety-critical system, which could lead to hazardous situations. To cope with this, we suggested a segmentation technique that makes hidden trigger backdoor attacks more robust. The tiny trigger patterns are conventionally established by a series of parameters encompassing their DNN size, location, color, shape, and other defining attributes. From the original triggers, alternate triggers are generated to control the backdoor patterns by a third party in addition to their original designer, which can produce a higher success rate than the original triggers. However, the significant downside of these approaches is the lack of automation in the scene segmentation phase, which results in the poor optimization of the threat model. We developed a novel technique that automatically generates alternate triggers to increase the effectiveness of triggers. Image denoising is performed for this purpose, followed by scene segmentation techniques to make the poisoned classifier more robust. The experimental results demonstrated that our proposed technique achieved 99% to 100% accuracy and helped reduce the vulnerabilities of DL models by exposing their loopholes." @default.
- W4362609169 created "2023-04-06" @default.
- W4362609169 creator A5000537746 @default.
- W4362609169 creator A5040374400 @default.
- W4362609169 creator A5064821311 @default.
- W4362609169 creator A5070647549 @default.
- W4362609169 creator A5073582886 @default.
- W4362609169 date "2023-04-05" @default.
- W4362609169 modified "2023-10-13" @default.
- W4362609169 title "Automated Segmentation to Make Hidden Trigger Backdoor Attacks Robust against Deep Neural Networks" @default.
- W4362609169 cites W2051267297 @default.
- W4362609169 cites W2535690855 @default.
- W4362609169 cites W2934843808 @default.
- W4362609169 cites W2942091739 @default.
- W4362609169 cites W3004478028 @default.
- W4362609169 cites W3034414373 @default.
- W4362609169 cites W3088837748 @default.
- W4362609169 cites W3098772125 @default.
- W4362609169 cites W3206937959 @default.
- W4362609169 cites W3206976809 @default.
- W4362609169 cites W3215147048 @default.
- W4362609169 cites W3217417806 @default.
- W4362609169 cites W4206244849 @default.
- W4362609169 cites W4297973329 @default.
- W4362609169 cites W4304945145 @default.
- W4362609169 cites W4312233756 @default.
- W4362609169 cites W4316654951 @default.
- W4362609169 cites W4317796293 @default.
- W4362609169 doi "https://doi.org/10.3390/app13074599" @default.
- W4362609169 hasPublicationYear "2023" @default.
- W4362609169 type Work @default.
- W4362609169 citedByCount "1" @default.
- W4362609169 crossrefType "journal-article" @default.
- W4362609169 hasAuthorship W4362609169A5000537746 @default.
- W4362609169 hasAuthorship W4362609169A5040374400 @default.
- W4362609169 hasAuthorship W4362609169A5064821311 @default.
- W4362609169 hasAuthorship W4362609169A5070647549 @default.
- W4362609169 hasAuthorship W4362609169A5073582886 @default.
- W4362609169 hasBestOaLocation W43626091691 @default.
- W4362609169 hasConcept C119857082 @default.
- W4362609169 hasConcept C153180895 @default.
- W4362609169 hasConcept C154945302 @default.
- W4362609169 hasConcept C2781045450 @default.
- W4362609169 hasConcept C2984842247 @default.
- W4362609169 hasConcept C31972630 @default.
- W4362609169 hasConcept C38652104 @default.
- W4362609169 hasConcept C41008148 @default.
- W4362609169 hasConcept C50644808 @default.
- W4362609169 hasConcept C89600930 @default.
- W4362609169 hasConceptScore W4362609169C119857082 @default.
- W4362609169 hasConceptScore W4362609169C153180895 @default.
- W4362609169 hasConceptScore W4362609169C154945302 @default.
- W4362609169 hasConceptScore W4362609169C2781045450 @default.
- W4362609169 hasConceptScore W4362609169C2984842247 @default.
- W4362609169 hasConceptScore W4362609169C31972630 @default.
- W4362609169 hasConceptScore W4362609169C38652104 @default.
- W4362609169 hasConceptScore W4362609169C41008148 @default.
- W4362609169 hasConceptScore W4362609169C50644808 @default.
- W4362609169 hasConceptScore W4362609169C89600930 @default.
- W4362609169 hasIssue "7" @default.
- W4362609169 hasLocation W43626091691 @default.
- W4362609169 hasOpenAccess W4362609169 @default.
- W4362609169 hasPrimaryLocation W43626091691 @default.
- W4362609169 hasRelatedWork W1669643531 @default.
- W4362609169 hasRelatedWork W1982826852 @default.
- W4362609169 hasRelatedWork W2005437358 @default.
- W4362609169 hasRelatedWork W2008656436 @default.
- W4362609169 hasRelatedWork W2023558673 @default.
- W4362609169 hasRelatedWork W2110230079 @default.
- W4362609169 hasRelatedWork W2134924024 @default.
- W4362609169 hasRelatedWork W2517104666 @default.
- W4362609169 hasRelatedWork W2613186388 @default.
- W4362609169 hasRelatedWork W1967061043 @default.
- W4362609169 hasVolume "13" @default.
- W4362609169 isParatext "false" @default.
- W4362609169 isRetracted "false" @default.
- W4362609169 workType "article" @default.