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) are energy efficient alternatives to commonlyused deep neural networks (DNNs). Through event-driven information processing,SNNs can reduce the expensive compute requirements of DNNs considerably, whileachieving comparable performance. However, high inference latency is asignificant hindrance to the edge deployment of deep SNNs. Computation overmultiple timesteps not only increases latency as well as overall energy budgetdue to higher number of operations, but also incurs memory access overhead offetching membrane potentials, both of which lessen the energy benefits of SNNs.To overcome this bottleneck and leverage the full potential of SNNs, we proposean Iterative Initialization and Retraining method for SNNs (IIR-SNN) to performsingle shot inference in the temporal axis. The method starts with an SNNtrained with T timesteps (T>1). Then at each stage of latency reduction, thenetwork trained at previous stage with higher timestep is utilized asinitialization for subsequent training with lower timestep. This acts as acompression method, as the network is gradually shrunk in the temporal domain.In this paper, we use direct input encoding and choose T=5, since as perliterature, it is the minimum required latency to achieve satisfactoryperformance on ImageNet. The proposed scheme allows us to obtain SNNs with upto unit latency, requiring a single forward pass during inference. We achievetop-1 accuracy of 93.05%, 70.15% and 67.71% on CIFAR-10, CIFAR-100 andImageNet, respectively using VGG16, with just 1 timestep. In addition, IIR-SNNsperform inference with 5-2500X reduced latency compared to otherstate-of-the-art SNNs, maintaining comparable or even better accuracy.Furthermore, in comparison with standard DNNs, the proposed IIR-SNNsprovide25-33X higher energy efficiency, while being comparable to them inclassification performance.

One Timestep is All You Need: Training Spiking Neural Networks with Ultra Low Latency