SpykeTorch: Efficient Simulation of Convolutional Spiking Neural Networks With at Most One Spike per Neuron

• Published in: Frontiers in neuroscience Vol. 13; p. 625
• Main Authors: Mozafari, Milad, Ganjtabesh, Mohammad, Nowzari-Dalini, Abbas, Masquelier, Timothée
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Research Foundation 12.07.2019; Frontiers; Frontiers Media S.A
• Subjects: Artificial intelligence; Cognitive science; Computer science; convolutional spiking neural networks; Deep learning; Energy; Firing pattern; Life Sciences; Neural networks; Neurons; Neurons and Cognition; Neuroscience; one spike per neuron; Reinforcement; reward-modulated STDP; STDP; tensor-based computing; time-to-first-spike coding; reward-modulated STDP; tensor-based computing; convolutional spiking neural networks; one spike per neuron; STDP; GPU acceleration; time-to-first-spike coding
• ISSN: 1662-453X; 1662-4548; 1662-453X
• DOI: 10.3389/fnins.2019.00625

Application of deep convolutional spiking neural networks (SNNs) to artificial intelligence (AI) tasks has recently gained a lot of interest since SNNs are hardware-friendly and energy-efficient. Unlike the non-spiking counterparts, most of the existing SNN simulation frameworks are not practically efficient enough for large-scale AI tasks. In this paper, we introduce SpykeTorch, an open-source high-speed simulation framework based on PyTorch. This framework simulates convolutional SNNs with at most one spike per neuron and the rank-order encoding scheme. In terms of learning rules, both spike-timing-dependent plasticity (STDP) and reward-modulated STDP (R-STDP) are implemented, but other rules could be implemented easily. Apart from the aforementioned properties, SpykeTorch is highly generic and capable of reproducing the results of various studies. Computations in the proposed framework are tensor-based and totally done by PyTorch functions, which in turn brings the ability of just-in-time optimization for running on CPUs, GPUs, or Multi-GPU platforms.