Exploring neural oscillations during speech perception via surrogate gradient spiking neural networks

arXiv:2404.14024v1 Announce Type: new
Abstract: Understanding cognitive processes in the brain demands sophisticated models capable of replicating neural dynamics at large scales. We present a physiologically inspired speech recognition architecture, compatible and scalable with deep learning frameworks, and demonstrate that end-to-end gradient descent training leads to the emergence of neural oscillations in the central spiking neural network. Significant cross-frequency couplings, indicative of these oscillations, are measured within and across network layers during speech processing, whereas no such interactions are observed …

abstract architecture arxiv brain cognitive cs.cl deep learning deep learning frameworks dynamics emergence frameworks gradient leads networks neural networks perception processes q-bio.nc recognition scalable speech speech recognition spiking neural networks training type understanding via