A network that performs brute-force conversion of a temporal sequence to a spatial pattern: relevance to odor recognition

Abstract

A classic problem in neuroscience is how temporal sequences can be recognized. This problem is exemplified in the olfactory system, where an odor is defined by the temporal sequence of olfactory bulb output that occurs during a sniff. This sequence is discrete because the output is subdivided by gamma frequency oscillations. Here we propose a new class of “brute-force” solutions to recognition of discrete sequences. We demonstrate a network architecture in which there are a small number of modules, each of which provides a persistent snapshot of what occurs in a different gamma cycle. The collection of these snapshots forms a spatial pattern that can be recognized by standard attractor-based network mechanisms. We will discuss the implications of this strategy for recognizing odor-specific sequences generated by the olfactory bulb.

Publication
Frontiers in Computational Neuroscience