Matches in SemOpenAlex for { <https://semopenalex.org/work/W3089934672> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W3089934672 abstract "Spiking Neural Networks (SNNs) offer a promising alternative to traditional deep learning frameworks, since they provide higher computational efficiency due to event-driven information processing. SNNs distribute the analog values of pixel intensities into binary spikes over time. However, the most widely used input coding schemes, such as Poisson based rate-coding, do not leverage the additional temporal learning capability of SNNs effectively. Moreover, these SNNs suffer from high inference latency which is a major bottleneck to their deployment. To overcome this, we propose a scalable time-based encoding scheme that utilizes the Discrete Cosine Transform (DCT) to reduce the number of timesteps required for inference. DCT decomposes an image into a weighted sum of sinusoidal basis images. At each time step, the Hadamard product of the DCT coefficients and a single frequency base, taken in order, is given to an accumulator that generates spikes upon crossing a threshold. We use the proposed scheme to learn DCT-SNN, a low-latency deep SNN with leaky-integrate-and-fire neurons, trained using surrogate gradient descent based backpropagation. We achieve top-1 accuracy of 89.94%, 68.3% and 52.43% on CIFAR-10, CIFAR-100 and TinyImageNet, respectively using VGG architectures. Notably, DCT-SNN performs inference with 2-14X reduced latency compared to other state-of-the-art SNNs, while achieving comparable accuracy to their standard deep learning counterparts. The dimension of the transform allows us to control the number of timesteps required for inference. Additionally, we can trade-off accuracy with latency in a principled manner by dropping the highest frequency components during inference." @default.
- W3089934672 created "2020-10-08" @default.
- W3089934672 creator A5031161187 @default.
- W3089934672 creator A5066407038 @default.
- W3089934672 creator A5081219427 @default.
- W3089934672 date "2020-10-05" @default.
- W3089934672 modified "2023-09-27" @default.
- W3089934672 title "DCT-SNN: Using DCT to Distribute Spatial Information over Time for Learning Low-Latency Spiking Neural Networks." @default.
- W3089934672 cites W1533861849 @default.
- W3089934672 cites W1645800954 @default.
- W3089934672 cites W1999085092 @default.
- W3089934672 cites W2020676607 @default.
- W3089934672 cites W2031614119 @default.
- W3089934672 cites W2140196014 @default.
- W3089934672 cites W2150355110 @default.
- W3089934672 cites W2160815625 @default.
- W3089934672 cites W2233731247 @default.
- W3089934672 cites W2541249978 @default.
- W3089934672 cites W2584959338 @default.
- W3089934672 cites W2775079417 @default.
- W3089934672 cites W2806066966 @default.
- W3089934672 cites W2891530223 @default.
- W3089934672 cites W2904620099 @default.
- W3089934672 cites W2914117107 @default.
- W3089934672 cites W2962908174 @default.
- W3089934672 cites W2963834742 @default.
- W3089934672 cites W2964296416 @default.
- W3089934672 cites W2964338223 @default.
- W3089934672 cites W2987748894 @default.
- W3089934672 cites W2990793844 @default.
- W3089934672 cites W2995289984 @default.
- W3089934672 cites W3007283957 @default.
- W3089934672 cites W3007926765 @default.
- W3089934672 cites W3013009725 @default.
- W3089934672 cites W3015205410 @default.
- W3089934672 cites W3016620475 @default.
- W3089934672 cites W3034771037 @default.
- W3089934672 cites W3102087395 @default.
- W3089934672 cites W3159754263 @default.
- W3089934672 cites W2006370340 @default.
- W3089934672 hasPublicationYear "2020" @default.
- W3089934672 type Work @default.
- W3089934672 sameAs 3089934672 @default.
- W3089934672 citedByCount "2" @default.
- W3089934672 countsByYear W30899346722021 @default.
- W3089934672 crossrefType "posted-content" @default.
- W3089934672 hasAuthorship W3089934672A5031161187 @default.
- W3089934672 hasAuthorship W3089934672A5066407038 @default.
- W3089934672 hasAuthorship W3089934672A5081219427 @default.
- W3089934672 hasConcept C108583219 @default.
- W3089934672 hasConcept C11413529 @default.
- W3089934672 hasConcept C115961682 @default.
- W3089934672 hasConcept C11731999 @default.
- W3089934672 hasConcept C153180895 @default.
- W3089934672 hasConcept C154945302 @default.
- W3089934672 hasConcept C2221639 @default.
- W3089934672 hasConcept C2776214188 @default.
- W3089934672 hasConcept C41008148 @default.
- W3089934672 hasConcept C50644808 @default.
- W3089934672 hasConceptScore W3089934672C108583219 @default.
- W3089934672 hasConceptScore W3089934672C11413529 @default.
- W3089934672 hasConceptScore W3089934672C115961682 @default.
- W3089934672 hasConceptScore W3089934672C11731999 @default.
- W3089934672 hasConceptScore W3089934672C153180895 @default.
- W3089934672 hasConceptScore W3089934672C154945302 @default.
- W3089934672 hasConceptScore W3089934672C2221639 @default.
- W3089934672 hasConceptScore W3089934672C2776214188 @default.
- W3089934672 hasConceptScore W3089934672C41008148 @default.
- W3089934672 hasConceptScore W3089934672C50644808 @default.
- W3089934672 hasLocation W30899346721 @default.
- W3089934672 hasOpenAccess W3089934672 @default.
- W3089934672 hasPrimaryLocation W30899346721 @default.
- W3089934672 hasRelatedWork W2343815575 @default.
- W3089934672 hasRelatedWork W2795627216 @default.
- W3089934672 hasRelatedWork W2891530223 @default.
- W3089934672 hasRelatedWork W2896611294 @default.
- W3089934672 hasRelatedWork W2897623300 @default.
- W3089934672 hasRelatedWork W2912945917 @default.
- W3089934672 hasRelatedWork W2989937740 @default.
- W3089934672 hasRelatedWork W3006664330 @default.
- W3089934672 hasRelatedWork W3022766987 @default.
- W3089934672 hasRelatedWork W3035008388 @default.
- W3089934672 hasRelatedWork W3039834782 @default.
- W3089934672 hasRelatedWork W3121127024 @default.
- W3089934672 hasRelatedWork W3161328488 @default.
- W3089934672 hasRelatedWork W3186531543 @default.
- W3089934672 hasRelatedWork W3188343682 @default.
- W3089934672 hasRelatedWork W3199957588 @default.
- W3089934672 hasRelatedWork W3201013180 @default.
- W3089934672 hasRelatedWork W3204436175 @default.
- W3089934672 hasRelatedWork W3206768834 @default.
- W3089934672 hasRelatedWork W3211294500 @default.
- W3089934672 isParatext "false" @default.
- W3089934672 isRetracted "false" @default.
- W3089934672 magId "3089934672" @default.
- W3089934672 workType "article" @default.