Gaspard Goupy

Gaspard Goupy

PhD student working on spiking neural networks

Unsupervised and efficient learning in sparsely activated convolutional spiking neural networks enabled by voltage-dependent synaptic plasticity

Published in Neuromorphic Computing and Engineering, 2023

Paper - Code

In this paper, we implement Voltage-Dependent Synaptic Plasticity (VDSP), an unsupervised and hardware-friendly local learning rule, in convolutional spiking neural networks.

Figure of the CSNN-VDSP architecture

Recommended citation: Gaspard Goupy, Alexandre Juneau-Fecteau, Nikhil Garg, Ismael Balafrej, Fabien Alibart, Luc Frechette, Dominique Drouin, and Yann Beilliard. Unsupervised and Efficient Learning in Sparsely Activated Convolutional Spiking Neural Networks Enabled by Voltage-Dependent Synaptic Plasticity. Neuromorphic Computing and Engineering, 3, 2023.