2023 BrainChip Discussion, page-1118

Overview of Spiking Neural Network Learning Approaches and Their Computational Complexities
by
Paweł Pietrzak
,
Szymon Szczęsny
*,
Damian Huderek
and
Łukasz Przyborowski



Institute of Computing Science, Faculty of Computing and Telecommunications, Poznan University of Technology, Piotrowo 3A Street, 61-138 Poznań, Poland
*
Author to whom correspondence should be addressed.
Sensors 2023, 23(6), 3037; https://doi.org/10.3390/s23063037
Received: 7 February 2023 / Revised: 8 March 2023 /Accepted: 9 March 2023 / Published: 11 March 2023

Currently existing neuromorphic architectures include

• IBM TrueNorth,
• Intel Loihi,
• Tianjic,
• SpiNNaker,
• BrainScaleS,
• NeuronFlow,
• DYNAP, and
• Akida.

Some of the above architectures are fully neuromorphic [31,32], while other remain hybrid, meaning that they use asynchronous circuits together with synchronous processors [33,34]. Despite the field being still in its infancy, the first commercial neuromorphic processor was made available worldwide in August 2021. It is Akida from Australian company BrainChip. Unfortunately, these hardware platforms are very expensive at the time of writing and (apart from Akida) not feasibly available.
Abstract
Spiking neural networks (SNNs) are subjects of a topic that is gaining more and more interest nowadays. They more closely resemble actual neural networks in the brain than their second-generation counterparts, artificial neural networks (ANNs). SNNs have the potential to be more energy efficient than ANNs on event-driven neuromorphic hardware. This can yield drastic maintenance cost reduction for neural network models, as the energy consumption would be much lower in comparison to regular deep learning models hosted in the cloud today. However, such hardware is still not yet widely available. On standard computer architectures consisting mainly of central processing units (CPUs) and graphics processing units (GPUs) ANNs, due to simpler models of neurons and simpler models of connections between neurons, have the upper hand in terms of execution speed. In general, they also win in terms of learning algorithms, as SNNs do not reach the same levels of performance as their second-generation counterparts in typical machine learning benchmark tasks, such as classification. In this paper, we review existing learning algorithms for spiking neural networks, divide them into categories by type, and assess their computational complexity.