Brain-inspired neural circuit evolution for spiking neural networks

Abstract

In biological neural systems, different neurons are capable of self-organizing to form different neural circuits for achieving a variety of cognitive functions. However, the current design paradigm of spiking neural networks is based on structures derived from deep learning. Such structures are dominated by feedforward connections without taking into account different types of neurons, which significantly prevent spiking neural networks from realizing their potential on complex tasks. It remains an open challenge to apply the rich dynamical properties of biological neural circuits to model the structure of current spiking neural networks. This paper provides a more biologically plausible evolutionary space by combining feedforward and feedback connections with excitatory and inhibitory neurons. We exploit the local spiking behavior of neurons to adaptively evolve neural circuits such as forward excitation, forward inhibition, feedback inhibition, and lateral inhibition by the local law of spike-timing-dependent plasticity and update the synaptic weights in combination with the global error signals. By using the evolved neural circuits, we construct spiking neural networks for image classification and reinforcement learning tasks. Using the brain-inspired Neural circuit Evolution strategy (NeuEvo) with rich neural circuit types, the evolved spiking neural network greatly enhances capability on perception and reinforcement learning tasks. NeuEvo achieves state-of-the-art performance on CIFAR10, DVS-CIFAR10, DVS-Gesture, and N-Caltech101 datasets and achieves advanced performance on ImageNet. Combined with on-policy and off-policy deep reinforcement learning algorithms, it achieves comparable performance with artificial neural networks. The evolved spiking neural circuits lay the foundation for the evolution of complex networks with functions.

Learning fast while changing slow in spiking neural networks

“…The results not only showcase the effectiveness of the proposed methodology but also highlight its potential impact on real-world applications. This work contributes to the ongoing discourse on the intersection of RL, computational neuroscience, and artificial intelligence, pushing the boundaries of what is achievable in the realm of autonomous, adaptive systems [12,31].…”

Editorial: Cutting-edge systems and materials for brain-inspired computing, adaptive bio-interfacing and smart sensing: implications for neuromorphic computing and biointegrated frameworks

“…Spiking neural networks (SNNs) are more competitive with non-spiking models in many computer vision tasks. Using a brain-inspired neural circuit evolution strategy and a rich set of neural circuit types, Shen et al ( 2023 ) evolved SNNs that greatly enhance perceptual and reinforcement learning tasks while combining online and offline deep reinforcement learning algorithms. Yu M. et al proposed a novel optimization algorithm for TTFS-based SNN systems that improved the accuracy of SNNs in terms of TTFS encoding and training.…”

“…Spiking neural networks (SNNs) are more competitive with non-spiking models in many computer vision tasks. Using a brain-inspired neural circuit evolution strategy and a rich set of neural circuit types, Shen et al (2023) Nevertheless, the SNN training algorithms are not mature enough for brain-inspired computing. Thus, realizing the conversion between SNN and ANN helps to improve the accuracy of brain-inspired computing.…”