Matches in SemOpenAlex for { <https://semopenalex.org/work/W4200038890> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W4200038890 abstract "Spiking Neural Networks (SNNs) are considered the third generation of NNs and can reach similar accuracy as conventional deep NNs, but with a considerable improvement in efficiency. However, to achieve high accuracy, state-of-the-art SNNs employ stochastic spike coding of the inputs, requiring multiple cycles of computation. Because of this and due to the nature of analog computing, it is required to accumulate and hold the charges of multiple cycles, necessitating a large membrane capacitor. This results in high energy, long latency, and expensive area costs, constituting one of the major bottlenecks in analog SNN implementations. Membrane capacitor size determines the precision of the firing time. Hence reducing the capacitor size considerably degrades the inference accuracy. To alleviate this, we focus on bridging the gap between binarized NNs (BNNs) and SNNs. BNNs are rapidly emerging as an attractive alternative for NNs due to their high efficiency and error tolerance. In this work, we evaluate the impact of deploying error-resilient BNNs, i.e. BNNs that have been proactively trained in the presence of errors, on analog implementation of SNNs. We show that for BNNs, the capacitor size and latency can be reduced significantly compared to state-of-the-art SNNs, which employ multi-bit models. Our experiments demonstrate that when error-resilient BNNs are deployed on analog-based SNN accelerator, the size of the membrane capacitor is reduced by 50%, the inference latency is decreased by two orders of magnitude, and energy is reduced by 57% compared to the baseline 4-bit SNN implementation, under minimal accuracy cost." @default.
- W4200038890 created "2021-12-31" @default.
- W4200038890 creator A5025566655 @default.
- W4200038890 creator A5046948316 @default.
- W4200038890 creator A5059133190 @default.
- W4200038890 creator A5067947688 @default.
- W4200038890 creator A5077475969 @default.
- W4200038890 creator A5089877293 @default.
- W4200038890 date "2021-11-01" @default.
- W4200038890 modified "2023-09-26" @default.
- W4200038890 title "Binarized SNNs: Efficient and Error-Resilient Spiking Neural Networks through Binarization" @default.
- W4200038890 cites W2140707343 @default.
- W4200038890 cites W2518281301 @default.
- W4200038890 cites W2771127423 @default.
- W4200038890 cites W2798654995 @default.
- W4200038890 cites W2912834930 @default.
- W4200038890 cites W2932447952 @default.
- W4200038890 cites W2946942415 @default.
- W4200038890 cites W2956713401 @default.
- W4200038890 cites W2963335874 @default.
- W4200038890 cites W2967507173 @default.
- W4200038890 cites W2988640543 @default.
- W4200038890 cites W3012119183 @default.
- W4200038890 cites W3022224321 @default.
- W4200038890 cites W3022517736 @default.
- W4200038890 cites W3024813878 @default.
- W4200038890 cites W3037354233 @default.
- W4200038890 cites W3038044676 @default.
- W4200038890 cites W3039362986 @default.
- W4200038890 cites W3091691040 @default.
- W4200038890 cites W3124237980 @default.
- W4200038890 cites W3135322505 @default.
- W4200038890 cites W3158218720 @default.
- W4200038890 cites W3158377181 @default.
- W4200038890 cites W3159974581 @default.
- W4200038890 cites W3184263515 @default.
- W4200038890 cites W3185026020 @default.
- W4200038890 cites W4251155475 @default.
- W4200038890 doi "https://doi.org/10.1109/iccad51958.2021.9643463" @default.
- W4200038890 hasPublicationYear "2021" @default.
- W4200038890 type Work @default.
- W4200038890 citedByCount "3" @default.
- W4200038890 countsByYear W42000388902021 @default.
- W4200038890 countsByYear W42000388902022 @default.
- W4200038890 countsByYear W42000388902023 @default.
- W4200038890 crossrefType "proceedings-article" @default.
- W4200038890 hasAuthorship W4200038890A5025566655 @default.
- W4200038890 hasAuthorship W4200038890A5046948316 @default.
- W4200038890 hasAuthorship W4200038890A5059133190 @default.
- W4200038890 hasAuthorship W4200038890A5067947688 @default.
- W4200038890 hasAuthorship W4200038890A5077475969 @default.
- W4200038890 hasAuthorship W4200038890A5089877293 @default.
- W4200038890 hasConcept C113775141 @default.
- W4200038890 hasConcept C11731999 @default.
- W4200038890 hasConcept C119599485 @default.
- W4200038890 hasConcept C127413603 @default.
- W4200038890 hasConcept C154945302 @default.
- W4200038890 hasConcept C2742236 @default.
- W4200038890 hasConcept C2776214188 @default.
- W4200038890 hasConcept C41008148 @default.
- W4200038890 hasConcept C50644808 @default.
- W4200038890 hasConcept C76155785 @default.
- W4200038890 hasConcept C82876162 @default.
- W4200038890 hasConceptScore W4200038890C113775141 @default.
- W4200038890 hasConceptScore W4200038890C11731999 @default.
- W4200038890 hasConceptScore W4200038890C119599485 @default.
- W4200038890 hasConceptScore W4200038890C127413603 @default.
- W4200038890 hasConceptScore W4200038890C154945302 @default.
- W4200038890 hasConceptScore W4200038890C2742236 @default.
- W4200038890 hasConceptScore W4200038890C2776214188 @default.
- W4200038890 hasConceptScore W4200038890C41008148 @default.
- W4200038890 hasConceptScore W4200038890C50644808 @default.
- W4200038890 hasConceptScore W4200038890C76155785 @default.
- W4200038890 hasConceptScore W4200038890C82876162 @default.
- W4200038890 hasLocation W42000388901 @default.
- W4200038890 hasOpenAccess W4200038890 @default.
- W4200038890 hasPrimaryLocation W42000388901 @default.
- W4200038890 hasRelatedWork W3157350647 @default.
- W4200038890 hasRelatedWork W3195610113 @default.
- W4200038890 hasRelatedWork W4200635499 @default.
- W4200038890 hasRelatedWork W4213041209 @default.
- W4200038890 hasRelatedWork W4285327616 @default.
- W4200038890 hasRelatedWork W4293023302 @default.
- W4200038890 hasRelatedWork W4300865491 @default.
- W4200038890 hasRelatedWork W4308075506 @default.
- W4200038890 hasRelatedWork W4323323083 @default.
- W4200038890 hasRelatedWork W4366342874 @default.
- W4200038890 isParatext "false" @default.
- W4200038890 isRetracted "false" @default.
- W4200038890 workType "article" @default.