Prof. Ed Large on Networks of Oscillators for Sound Perception

Signal Processing, Plasticity and Pattern Formation in Networks of Neural Oscillators

ABSTRACT: I will introduce a new framework for nonlinear signal processing called Gradient Frequency Neural Networks (GrFNNs). GrFNNs are networks of active nonlinear elements that mimic small populations of neurons in the brain. Networks self-organize when stimulated with sound, and they can learn to recognize patterns. GrFNNs are based on a canonical model of neural oscillation, modeling an entire family of dynamical systems. We have recently released the GrFNN Toolbox, a software library that allows the simulation of signal processing, pattern formation and Hebbian plasticity. I will give examples of how GrFNNs can be used to model cochlea, brainstem, and auditory cortical responses to sound, with applications to pitch perception, consonance, musical tonality, and rhythm. 

BIO: Ed Large directs the Music Dynamics Laboratory at University of Connecticut, where he is a Professor of Psychology and Professor of Physics. Ed received his Ph.D. from The Ohio State University in 1994, and he did his postdoctoral work at University of Pennsylvania. His research areas include music psychology, auditory neuroscience and nonlinear dynamical systems. He uses theoretical modeling in conjunction with behavioral, neurophysiological and neuroimaging techniques to understand how people respond to complex, temporally structured sequences of sound such as music and speech. He and his colleagues have pioneered the idea that attention is a dynamic, and inherently rhythmic process. He has applied these ideas to explain the rhythmic structure of music, and its interaction with brain dynamics. His current research projects include auditory pattern recognition and learning, perception of tonality in music, auditory brainstem neurodynamics, cortical dynamics of attention, perception of rhythm in music and speech, rhythmic interactions in nonhuman primates and emotional communication in music. He currently serves as President of the Society for Music Perception and Cognition.