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) well support spatiotemporal learning andenergy-efficient event-driven hardware neuromorphic processors. As an importantclass of SNNs, recurrent spiking neural networks (RSNNs) possess greatcomputational power. However, the practical application of RSNNs is severelylimited by challenges in training. Biologically-inspired unsupervised learninghas limited capability in boosting the performance of RSNNs. On the other hand,existing backpropagation (BP) methods suffer from high complexity of unrollingin time, vanishing and exploding gradients, and approximate differentiation ofdiscontinuous spiking activities when applied to RSNNs. To enable supervisedtraining of RSNNs under a well-defined loss function, we present a novelSpike-Train level RSNNs Backpropagation (ST-RSBP) algorithm for training deepRSNNs. The proposed ST-RSBP directly computes the gradient of a rated-codedloss function defined at the output layer of the network w.r.t tunableparameters. The scalability of ST-RSBP is achieved by the proposed spike-trainlevel computation during which temporal effects of the SNN is captured in boththe forward and backward pass of BP. Our ST-RSBP algorithm can be broadlyapplied to RSNNs with a single recurrent layer or deep RSNNs with multiplefeed-forward and recurrent layers. Based upon challenging speech and imagedatasets including TI46, N-TIDIGITS, Fashion-MNIST and MNIST, ST-RSBP is ableto train RSNNs with an accuracy surpassing that of the current state-of-art SNNBP algorithms and conventional non-spiking deep learning models.

Spike-Train Level Backpropagation for Training Deep Recurrent Spiking Neural Networks