Matches in SemOpenAlex for { <https://semopenalex.org/work/W3187007027> ?p ?o ?g. }
Showing items 1 to 93 of
93
with 100 items per page.
- W3187007027 abstract "Deep Learning (DL) algorithms have gained popularity owing to their practical problem-solving capacity. However, they suffer from a serious integrity threat, i.e., their vulnerability to adversarial attacks. In the quest for DL trustworthiness, recent works claimed the inherent robustness of Spiking Neural Networks (SNNs) to these attacks, without considering the variability in their structural spiking parameters. This paper explores the security enhancement of SNNs through internal structural parameters. Specifically, we investigate the SNNs robustness to adversarial attacks with different values of the neuron's firing voltage thresholds and time window boundaries. We thoroughly study SNNs security under different adversarial attacks in the strong white-box setting, with different noise budgets and under variable spiking parameters. Our results show a significant impact of the structural parameters on the SNNs' security, and promising sweet spots can be reached to design trustworthy SNNs with 85% higher robustness than a traditional non-spiking DL system. To the best of our knowledge, this is the first work that investigates the impact of structural parameters on SNNs robustness to adversarial attacks. The proposed contributions and the experimental framework is available online <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup> <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup> https://github.com/rda-ela/SNN-Adversarial-Attacks to the community for reproducible research." @default.
- W3187007027 created "2021-08-02" @default.
- W3187007027 creator A5002123043 @default.
- W3187007027 creator A5042642081 @default.
- W3187007027 creator A5057618925 @default.
- W3187007027 creator A5065473138 @default.
- W3187007027 date "2021-02-01" @default.
- W3187007027 modified "2023-10-14" @default.
- W3187007027 title "Securing Deep Spiking Neural Networks against Adversarial Attacks through Inherent Structural Parameters" @default.
- W3187007027 cites W101771737 @default.
- W3187007027 cites W2108069432 @default.
- W3187007027 cites W2583832915 @default.
- W3187007027 cites W2783525259 @default.
- W3187007027 cites W2899111971 @default.
- W3187007027 cites W2945789436 @default.
- W3187007027 cites W2947683811 @default.
- W3187007027 cites W2963271936 @default.
- W3187007027 cites W2978279179 @default.
- W3187007027 cites W2987441303 @default.
- W3187007027 cites W3003366625 @default.
- W3187007027 cites W3038988173 @default.
- W3187007027 cites W3090384356 @default.
- W3187007027 cites W3091431973 @default.
- W3187007027 cites W3091443325 @default.
- W3187007027 cites W3091691040 @default.
- W3187007027 cites W3105904398 @default.
- W3187007027 cites W3109080038 @default.
- W3187007027 cites W3120496366 @default.
- W3187007027 doi "https://doi.org/10.23919/date51398.2021.9473981" @default.
- W3187007027 hasPublicationYear "2021" @default.
- W3187007027 type Work @default.
- W3187007027 sameAs 3187007027 @default.
- W3187007027 citedByCount "11" @default.
- W3187007027 countsByYear W31870070272021 @default.
- W3187007027 countsByYear W31870070272022 @default.
- W3187007027 countsByYear W31870070272023 @default.
- W3187007027 crossrefType "proceedings-article" @default.
- W3187007027 hasAuthorship W3187007027A5002123043 @default.
- W3187007027 hasAuthorship W3187007027A5042642081 @default.
- W3187007027 hasAuthorship W3187007027A5057618925 @default.
- W3187007027 hasAuthorship W3187007027A5065473138 @default.
- W3187007027 hasBestOaLocation W31870070272 @default.
- W3187007027 hasConcept C104317684 @default.
- W3187007027 hasConcept C108583219 @default.
- W3187007027 hasConcept C11731999 @default.
- W3187007027 hasConcept C119857082 @default.
- W3187007027 hasConcept C153180895 @default.
- W3187007027 hasConcept C153701036 @default.
- W3187007027 hasConcept C154945302 @default.
- W3187007027 hasConcept C185592680 @default.
- W3187007027 hasConcept C2984842247 @default.
- W3187007027 hasConcept C37736160 @default.
- W3187007027 hasConcept C38652104 @default.
- W3187007027 hasConcept C41008148 @default.
- W3187007027 hasConcept C50644808 @default.
- W3187007027 hasConcept C55493867 @default.
- W3187007027 hasConcept C63479239 @default.
- W3187007027 hasConceptScore W3187007027C104317684 @default.
- W3187007027 hasConceptScore W3187007027C108583219 @default.
- W3187007027 hasConceptScore W3187007027C11731999 @default.
- W3187007027 hasConceptScore W3187007027C119857082 @default.
- W3187007027 hasConceptScore W3187007027C153180895 @default.
- W3187007027 hasConceptScore W3187007027C153701036 @default.
- W3187007027 hasConceptScore W3187007027C154945302 @default.
- W3187007027 hasConceptScore W3187007027C185592680 @default.
- W3187007027 hasConceptScore W3187007027C2984842247 @default.
- W3187007027 hasConceptScore W3187007027C37736160 @default.
- W3187007027 hasConceptScore W3187007027C38652104 @default.
- W3187007027 hasConceptScore W3187007027C41008148 @default.
- W3187007027 hasConceptScore W3187007027C50644808 @default.
- W3187007027 hasConceptScore W3187007027C55493867 @default.
- W3187007027 hasConceptScore W3187007027C63479239 @default.
- W3187007027 hasLocation W31870070271 @default.
- W3187007027 hasLocation W31870070272 @default.
- W3187007027 hasLocation W31870070273 @default.
- W3187007027 hasLocation W31870070274 @default.
- W3187007027 hasLocation W31870070275 @default.
- W3187007027 hasOpenAccess W3187007027 @default.
- W3187007027 hasPrimaryLocation W31870070271 @default.
- W3187007027 hasRelatedWork W2950183588 @default.
- W3187007027 hasRelatedWork W3193857078 @default.
- W3187007027 hasRelatedWork W3208304128 @default.
- W3187007027 hasRelatedWork W3208723233 @default.
- W3187007027 hasRelatedWork W4311734044 @default.
- W3187007027 hasRelatedWork W4320076403 @default.
- W3187007027 hasRelatedWork W4322759769 @default.
- W3187007027 hasRelatedWork W4379255972 @default.
- W3187007027 hasRelatedWork W4383955378 @default.
- W3187007027 hasRelatedWork W4286890323 @default.
- W3187007027 isParatext "false" @default.
- W3187007027 isRetracted "false" @default.
- W3187007027 magId "3187007027" @default.
- W3187007027 workType "article" @default.