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

In this work, we propose ReStoCNet, a residual stochastic multilayerconvolutional Spiking Neural Network (SNN) composed of binary kernels, toreduce the synaptic memory footprint and enhance the computational efficiencyof SNNs for complex pattern recognition tasks. ReStoCNet consists of an inputlayer followed by stacked convolutional layers for hierarchical input featureextraction, pooling layers for dimensionality reduction, and fully-connectedlayer for inference. In addition, we introduce residual connections between thestacked convolutional layers to improve the hierarchical feature learningcapability of deep SNNs. We propose Spike Timing Dependent Plasticity (STDP)based probabilistic learning algorithm, referred to as Hybrid-STDP (HB-STDP),incorporating Hebbian and anti-Hebbian learning mechanisms, to train the binarykernels forming ReStoCNet in a layer-wise unsupervised manner. We demonstratethe efficacy of ReStoCNet and the presented HB-STDP based unsupervised trainingmethodology on the MNIST and CIFAR-10 datasets. We show that residualconnections enable the deeper convolutional layers to self-learn usefulhigh-level input features and mitigate the accuracy loss observed in deep SNNsdevoid of residual connections. The proposed ReStoCNet offers >20x kernelmemory compression compared to full-precision (32-bit) SNN while yielding highenough classification accuracy on the chosen pattern recognition tasks.

ReStoCNet: Residual Stochastic Binary Convolutional Spiking Neural Network for Memory-Efficient Neuromorphic Computing