English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Spiking neural networks (SNNs) can utilize spatio-temporal information andhave a nature of energy efficiency which is a good alternative to deep neuralnetworks(DNNs). The event-driven information processing makes SNNs can reducethe expensive computation of DNNs and save a lot of energy consumption.However, high training and inference latency is a limitation of the developmentof deeper SNNs. SNNs usually need tens or even hundreds of time steps duringthe training and inference process which causes not only the increase oflatency but also the waste of energy consumption. To overcome this problem, weproposed a novel training method based on backpropagation (BP) for ultra-lowlatency(1-2 time steps) SNN with multi-threshold. In order to increase theinformation capacity of each spike, we introduce the multi-threshold LeakyIntegrate and Fired (LIF) model. In our proposed training method, we proposedthree approximated derivative for spike activity to solve the problem of thenon-differentiable issue which cause difficulties for direct training SNNsbased on BP. The experimental results show that our proposed method achieves anaverage accuracy of 99.56%, 93.08%, and 87.90% on MNIST, FashionMNIST, andCIFAR10, respectively with only 2 time steps. For the CIFAR10 dataset, ourproposed method achieve 1.12% accuracy improvement over the previously reporteddirect trained SNNs with fewer time steps.

Direct Training via Backpropagation for Ultra-low Latency Spiking Neural Networks with Multi-threshold