An Analog VLSI Model of Periodicity Extraction

Schaik, André van

Neural Information Processing Systems 

This paper presents an electronic system that extracts the periodicity of a sound. It uses three analogue VLSI building blocks: a silicon cochlea, two inner-hair-cell circuits and two spiking neuron chips. The silicon cochlea consists of a cascade of filters. Because of the delay between two outputs from the silicon cochlea, spike trains created at these outputs are synchronous only for a narrow range of periodicities. In contrast to traditional bandpass filters,where an increase in' selectivity has to be traded off against a decrease in response time, the proposed system responds quickly, independent of selectivity. 1 Introduction The human ear transduces airborne sounds into a neural signal using three stages in the inner ear's cochlea: (i) the mechanical filtering of the Basilar Membrane (BM), (ii) the transduction of membrane vibration into neurotransmitter release by the Inner Hair Cells (IHCs), and (iii) spike generation by the Spiral Ganglion Cells (SGCs), whose axons form the auditory nerve.