Odour Recognition and Segmentation by a model Olfactory Bulb and Cortex

Zhaoping Li, Gatsby Computational Neuroscience Unit, UCL
John Hertz, Nordita, Denmark

GCNU TR 1999-004 [October 1999]

Abstract
We present a model of an olfactory system that performs odor segmentation. Based on the anatomy and physiology of natural olfactory systems, it consists of a pain of coupled modules, bulb and cortex.  The bulb encodes the odor inputs as oscillating patterns.   The cortex functions as an associative memory:  When the input from the bulb matches a pattern stored in the connections between its units, the cortical units resonate in an oscillatory pattern characteristic of that odor.  Further circuitry transforms this oscillatory signal to a slowly-varying feedback to the bulb.  This feedback implements olfactory segmentation by suppressing the bulbar response to the pre-existing odor, thereby allowing subsequent odors to be singled out for recognition.

---

Download:  ps.gz  or  pdf