Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385484718> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W4385484718 abstract "Spiking neural networks (SNN) are able to learn spatiotemporal features while using less energy, especially on neuromorphic hardware. The most widely used spiking neuron in deep learning is the Leaky Integrate and Fire (LIF) neuron. LIF neurons operate sequentially, however, since the computation of state at time <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$t$</tex> relies on the state at time <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$t-1$</tex> being computed. This limitation is shared with Recurrent Neural Networks (RNN) and results in slow training on Graphics Processing Units (GPU). In this paper, we propose the Stochastic Parallelizable Spiking Neuron (SPSN) to overcome the sequential training limitation of LIF neurons. By separating the linear integration component from the non-linear spiking function, SPSN can be run in parallel over time. The proposed approach results in performance comparable with the state-of-the-art for feedforward neural networks on the Spiking Heidelberg Digits (SHD) dataset, outperforming LIF networks while training 10 times faster and outperforming non-spiking networks with the same network architecture. For longer input sequences of 10 000 time-steps, we show that the proposed approach results in 4000 times faster training, thus demonstrating the potential of the proposed approach to accelerate SNN training for very large datasets." @default.
- W4385484718 created "2023-08-03" @default.
- W4385484718 creator A5016964600 @default.
- W4385484718 creator A5086006042 @default.
- W4385484718 date "2023-06-18" @default.
- W4385484718 modified "2023-09-26" @default.
- W4385484718 title "Accelerating SNN Training with Stochastic Parallelizable Spiking Neurons" @default.
- W4385484718 cites W2000335122 @default.
- W4385484718 cites W2051137788 @default.
- W4385484718 cites W2073813638 @default.
- W4385484718 cites W2108598243 @default.
- W4385484718 cites W2109596721 @default.
- W4385484718 cites W2116017529 @default.
- W4385484718 cites W2130459697 @default.
- W4385484718 cites W2145431022 @default.
- W4385484718 cites W2149176744 @default.
- W4385484718 cites W2152014477 @default.
- W4385484718 cites W2154439184 @default.
- W4385484718 cites W2157239334 @default.
- W4385484718 cites W2159951683 @default.
- W4385484718 cites W2229595996 @default.
- W4385484718 cites W2606722458 @default.
- W4385484718 cites W2775855651 @default.
- W4385484718 cites W2783525259 @default.
- W4385484718 cites W2798878556 @default.
- W4385484718 cites W2802745787 @default.
- W4385484718 cites W2969184347 @default.
- W4385484718 cites W2978816734 @default.
- W4385484718 cites W2984844508 @default.
- W4385484718 cites W3046044791 @default.
- W4385484718 cites W3105037130 @default.
- W4385484718 cites W4231081240 @default.
- W4385484718 cites W4250981202 @default.
- W4385484718 cites W4312117455 @default.
- W4385484718 cites W4379528681 @default.
- W4385484718 doi "https://doi.org/10.1109/ijcnn54540.2023.10191884" @default.
- W4385484718 hasPublicationYear "2023" @default.
- W4385484718 type Work @default.
- W4385484718 citedByCount "0" @default.
- W4385484718 crossrefType "proceedings-article" @default.
- W4385484718 hasAuthorship W4385484718A5016964600 @default.
- W4385484718 hasAuthorship W4385484718A5086006042 @default.
- W4385484718 hasBestOaLocation W43854847182 @default.
- W4385484718 hasConcept C11413529 @default.
- W4385484718 hasConcept C11731999 @default.
- W4385484718 hasConcept C148047603 @default.
- W4385484718 hasConcept C151927369 @default.
- W4385484718 hasConcept C154945302 @default.
- W4385484718 hasConcept C41008148 @default.
- W4385484718 hasConcept C50644808 @default.
- W4385484718 hasConceptScore W4385484718C11413529 @default.
- W4385484718 hasConceptScore W4385484718C11731999 @default.
- W4385484718 hasConceptScore W4385484718C148047603 @default.
- W4385484718 hasConceptScore W4385484718C151927369 @default.
- W4385484718 hasConceptScore W4385484718C154945302 @default.
- W4385484718 hasConceptScore W4385484718C41008148 @default.
- W4385484718 hasConceptScore W4385484718C50644808 @default.
- W4385484718 hasLocation W43854847181 @default.
- W4385484718 hasLocation W43854847182 @default.
- W4385484718 hasOpenAccess W4385484718 @default.
- W4385484718 hasPrimaryLocation W43854847181 @default.
- W4385484718 hasRelatedWork W2756276189 @default.
- W4385484718 hasRelatedWork W3089892344 @default.
- W4385484718 hasRelatedWork W3094559791 @default.
- W4385484718 hasRelatedWork W3117246098 @default.
- W4385484718 hasRelatedWork W4287184800 @default.
- W4385484718 hasRelatedWork W4306160827 @default.
- W4385484718 hasRelatedWork W4381827072 @default.
- W4385484718 hasRelatedWork W4385484718 @default.
- W4385484718 hasRelatedWork W4386227043 @default.
- W4385484718 hasRelatedWork W4386536395 @default.
- W4385484718 isParatext "false" @default.
- W4385484718 isRetracted "false" @default.
- W4385484718 workType "article" @default.