EqSpike: Spike-driven equilibrium propagation for neuromorphic implementations

Finding spike-based learning algorithms that can be implemented within the local constraints of neuromorphic systems, while achieving high accuracy, remains a formidable challenge. Equilibrium propagation is a promising alternative to backpropagation as it only involves local computations, but hardw...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:iScience Jg. 24; H. 3; S. 102222
Hauptverfasser: Martin, Erwann, Ernoult, Maxence, Laydevant, Jérémie, Li, Shuai, Querlioz, Damien, Petrisor, Teodora, Grollier, Julie
Format: Journal Article
Sprache:Englisch
Veröffentlicht: United States Elsevier Inc 19.03.2021
Elsevier
Schlagworte:
Algorithms
Artificial Intelligence
Computer Science
Physics
Artificial Intelligence
Algorithms
Computer Science
ISSN:2589-0042, 2589-0042
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Finding spike-based learning algorithms that can be implemented within the local constraints of neuromorphic systems, while achieving high accuracy, remains a formidable challenge. Equilibrium propagation is a promising alternative to backpropagation as it only involves local computations, but hardware-oriented studies have so far focused on rate-based networks. In this work, we develop a spiking neural network algorithm called EqSpike, compatible with neuromorphic systems, which learns by equilibrium propagation. Through simulations, we obtain a test recognition accuracy of 97.6% on the MNIST handwritten digits dataset (Mixed National Institute of Standards and Technology), similar to rate-based equilibrium propagation, and comparing favorably to alternative learning techniques for spiking neural networks. We show that EqSpike implemented in silicon neuromorphic technology could reduce the energy consumption of inference and training, respectively, by three orders and two orders of magnitude compared to graphics processing units. Finally, we also show that during learning, EqSpike weight updates exhibit a form of spike-timing-dependent plasticity, highlighting a possible connection with biology. [Display omitted] •EqSpike is a spiking neural network version of equilibrium propagation•It achieves 97.6% test accuracy on MNIST with a fully connected architecture•Its two-factor local learning rule is compatible with neuromorphic hardware•Its weight updates exhibit a form of spike-timing-dependent plasticity Computer Science; Algorithms; Artificial Intelligence
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
PMCID: PMC7970361
ISSN:2589-0042
2589-0042
DOI:10.1016/j.isci.2021.102222