Spiking Generative Adversarial Network with Attention Scoring Decoding

TL;DR

This paper introduces a novel spiking generative adversarial network with attention scoring decoding, addressing key challenges and demonstrating superior performance on multiple image datasets, aligning more closely with biological neural processing.

Contribution

It pioneers the development of a spiking GAN with attention-based decoding, improving image generation quality and biological plausibility compared to existing models.

Findings

Outperforms existing methods on MNIST, FashionMNIST, CIFAR10, CelebA datasets.

Addresses out-of-domain and temporal inconsistencies in spiking GANs.

Shows closer alignment with mouse neural processing patterns.

Abstract

Generative models based on neural networks present a substantial challenge within deep learning. As it stands, such models are primarily limited to the domain of artificial neural networks. Spiking neural networks, as the third generation of neural networks, offer a closer approximation to brain-like processing due to their rich spatiotemporal dynamics. However, generative models based on spiking neural networks are not well studied. In this work, we pioneer constructing a spiking generative adversarial network capable of handling complex images. Our first task was to identify the problems of out-of-domain inconsistency and temporal inconsistency inherent in spiking generative adversarial networks. We addressed these issues by incorporating the Earth-Mover distance and an attention-based weighted decoding method, significantly enhancing the performance of our algorithm across several datasets. Experimental results reveal that our approach outperforms existing methods on the MNIST, FashionMNIST, CIFAR10, and CelebA datasets. Moreover, compared with hybrid spiking generative adversarial networks, where the discriminator is an artificial analog neural network, our methodology demonstrates closer alignment with the information processing patterns observed in the mouse.