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) have great potential for energy-efficientimplementation of Deep Neural Networks (DNNs) on dedicated neuromorphichardware. Recent studies demonstrated competitive performance of SNNs comparedwith DNNs on image classification tasks, including CIFAR-10 and ImageNet data.The present work focuses on using SNNs in combination with deep reinforcementlearning in ATARI games, which involves additional complexity as compared toimage classification. We review the theory of converting DNNs to SNNs andextending the conversion to Deep Q-Networks (DQNs). We propose a robustrepresentation of the firing rate to reduce the error during the conversionprocess. In addition, we introduce a new metric to evaluate the conversionprocess by comparing the decisions made by the DQN and SNN, respectively. Wealso analyze how the simulation time and parameter normalization influence theperformance of converted SNNs. We achieve competitive scores on 17top-performing Atari games. To the best of our knowledge, our work is the firstto achieve state-of-the-art performance on multiple Atari games with SNNs. Ourwork serves as a benchmark for the conversion of DQNs to SNNs and paves the wayfor further research on solving reinforcement learning tasks with SNNs.

Strategy and Benchmark for Converting Deep Q-Networks to Event-Driven Spiking Neural Networks