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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Spiking neural networks (SNNs) offer an inherent ability to processspatial-temporal data, or in other words, realworld sensory data, but sufferfrom the difficulty of training high accuracy models. A major thread ofresearch on SNNs is on converting a pre-trained convolutional neural network(CNN) to an SNN of the same structure. State-of-the-art conversion methods areapproaching the accuracy limit, i.e., the near-zero accuracy loss of SNNagainst the original CNN. However, we note that this is made possible only whensignificantly more energy is consumed to process an input. In this paper, weargue that this trend of "energy for accuracy" is not necessary -- a littleenergy can go a long way to achieve the near-zero accuracy loss. Specifically,we propose a novel CNN-to-SNN conversion method that is able to use areasonably short spike train (e.g., 256 timesteps for CIFAR10 images) toachieve the near-zero accuracy loss. The new conversion method, named asexplicit current control (ECC), contains three techniques (currentnormalisation, thresholding for residual elimination, and consistencymaintenance for batch-normalisation), in order to explicitly control thecurrents flowing through the SNN when processing inputs. We implement ECC intoa tool nicknamed SpKeras, which can conveniently import Keras CNN models andconvert them into SNNs. We conduct an extensive set of experiments with thetool -- working with VGG16 and various datasets such as CIFAR10 and CIFAR100 --and compare with state-of-the-art conversion methods. Results show that ECC isa promising method that can optimise over energy consumption and accuracy losssimultaneously.

A Little Energy Goes a Long Way: Build an Energy-Efficient, Accurate Spiking Neural Network from Convolutional Neural Network