Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387397651> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W4387397651 endingPage "462" @default.
- W4387397651 startingPage "433" @default.
- W4387397651 abstract "Deep Neural Networks (DNNs) have achieved superhuman-like performance in several real-world applications such as classification, segmentation among others. Recently, analog crossbar architectures have been proposed as a viable in-memory computing alternative to improve the compute efficiency of DNNs for low power embedded applications. Although DNNs have achieved high performance, recent works have shown that they are vulnerable to adversarial attacks where small, imperceptible noise added to the input data can degrade the DNN performance. To this end, prior works have proposed algorithmic strategies such as adversarial classification and detection to mitigate the effect of adversarial attacks. However, these approaches are not energy-efficient and suffer from performance degradation when naively implemented on analog crossbars having various non-idealities inherently. To this end, in this chapter, we highlight efficiency-driven analog crossbar-aware approaches to improve the robustness of DNNs in two broad aspects: (1) Improving adversarial robustness of crossbar-mapped DNNs wherein, we discuss two recent works—NEAT and DetectX. (2) Examining and improving the natural robustness of crossbar-mapped structure-pruned DNN models against non-idealities." @default.
- W4387397651 created "2023-10-07" @default.
- W4387397651 creator A5004629816 @default.
- W4387397651 creator A5050310538 @default.
- W4387397651 creator A5050796355 @default.
- W4387397651 date "2023-10-07" @default.
- W4387397651 modified "2023-10-07" @default.
- W4387397651 title "Robustness for Embedded Machine Learning Using In-Memory Computing" @default.
- W4387397651 cites W2004823737 @default.
- W4387397651 cites W2056507634 @default.
- W4387397651 cites W2064814573 @default.
- W4387397651 cites W2307193480 @default.
- W4387397651 cites W2518281301 @default.
- W4387397651 cites W2612573399 @default.
- W4387397651 cites W2624863671 @default.
- W4387397651 cites W2740220207 @default.
- W4387397651 cites W2766179883 @default.
- W4387397651 cites W2782046614 @default.
- W4387397651 cites W2883149906 @default.
- W4387397651 cites W2904299207 @default.
- W4387397651 cites W2910506572 @default.
- W4387397651 cites W2913104037 @default.
- W4387397651 cites W2946522000 @default.
- W4387397651 cites W2951055820 @default.
- W4387397651 cites W2962933288 @default.
- W4387397651 cites W2963485691 @default.
- W4387397651 cites W2966199719 @default.
- W4387397651 cites W3010691265 @default.
- W4387397651 cites W3033076639 @default.
- W4387397651 cites W3033519076 @default.
- W4387397651 cites W3048300167 @default.
- W4387397651 cites W3091835145 @default.
- W4387397651 cites W3091922395 @default.
- W4387397651 cites W3092585568 @default.
- W4387397651 cites W3119147336 @default.
- W4387397651 cites W3142548120 @default.
- W4387397651 cites W3183458034 @default.
- W4387397651 cites W3198405160 @default.
- W4387397651 cites W3201417376 @default.
- W4387397651 cites W4236709213 @default.
- W4387397651 cites W4245199738 @default.
- W4387397651 cites W4245731639 @default.
- W4387397651 cites W4251775051 @default.
- W4387397651 cites W4280587271 @default.
- W4387397651 doi "https://doi.org/10.1007/978-3-031-40677-5_17" @default.
- W4387397651 hasPublicationYear "2023" @default.
- W4387397651 type Work @default.
- W4387397651 citedByCount "0" @default.
- W4387397651 crossrefType "book-chapter" @default.
- W4387397651 hasAuthorship W4387397651A5004629816 @default.
- W4387397651 hasAuthorship W4387397651A5050310538 @default.
- W4387397651 hasAuthorship W4387397651A5050796355 @default.
- W4387397651 hasConcept C104317684 @default.
- W4387397651 hasConcept C108583219 @default.
- W4387397651 hasConcept C113775141 @default.
- W4387397651 hasConcept C119857082 @default.
- W4387397651 hasConcept C154945302 @default.
- W4387397651 hasConcept C185592680 @default.
- W4387397651 hasConcept C2984842247 @default.
- W4387397651 hasConcept C29984679 @default.
- W4387397651 hasConcept C37736160 @default.
- W4387397651 hasConcept C41008148 @default.
- W4387397651 hasConcept C55493867 @default.
- W4387397651 hasConcept C63479239 @default.
- W4387397651 hasConcept C76155785 @default.
- W4387397651 hasConceptScore W4387397651C104317684 @default.
- W4387397651 hasConceptScore W4387397651C108583219 @default.
- W4387397651 hasConceptScore W4387397651C113775141 @default.
- W4387397651 hasConceptScore W4387397651C119857082 @default.
- W4387397651 hasConceptScore W4387397651C154945302 @default.
- W4387397651 hasConceptScore W4387397651C185592680 @default.
- W4387397651 hasConceptScore W4387397651C2984842247 @default.
- W4387397651 hasConceptScore W4387397651C29984679 @default.
- W4387397651 hasConceptScore W4387397651C37736160 @default.
- W4387397651 hasConceptScore W4387397651C41008148 @default.
- W4387397651 hasConceptScore W4387397651C55493867 @default.
- W4387397651 hasConceptScore W4387397651C63479239 @default.
- W4387397651 hasConceptScore W4387397651C76155785 @default.
- W4387397651 hasLocation W43873976511 @default.
- W4387397651 hasOpenAccess W4387397651 @default.
- W4387397651 hasPrimaryLocation W43873976511 @default.
- W4387397651 hasRelatedWork W2738001131 @default.
- W4387397651 hasRelatedWork W2950183588 @default.
- W4387397651 hasRelatedWork W2997056298 @default.
- W4387397651 hasRelatedWork W3080754722 @default.
- W4387397651 hasRelatedWork W3093978547 @default.
- W4387397651 hasRelatedWork W3127875750 @default.
- W4387397651 hasRelatedWork W3203790781 @default.
- W4387397651 hasRelatedWork W4285785480 @default.
- W4387397651 hasRelatedWork W4383221314 @default.
- W4387397651 hasRelatedWork W4386850404 @default.
- W4387397651 isParatext "false" @default.
- W4387397651 isRetracted "false" @default.
- W4387397651 workType "book-chapter" @default.