Perceptual Decision Making “Through the Eyes” of a Large-Scale Neural Model of V1

Jianing V Shi,Jim Wielaard,R Theodore Smith,Paul Sajda

doi:10.3389/fpsyg.2013.00161

Abstract

Sparse coding has been posited as an efficient information processing strategy employed by sensory systems, particularly visual cortex. Substantial theoretical and experimental work has focused on the issue of sparse encoding, namely how the early visual system maps the scene into a sparse representation. In this paper we investigate the complementary issue of sparse decoding, for example given activity generated by a realistic mapping of the visual scene to neuronal spike trains, how do downstream neurons best utilize this representation to generate a “decision.” Specifically we consider both sparse (L1-regularized) and non-sparse (L2 regularized) linear decoding for mapping the neural dynamics of a large-scale spiking neuron model of primary visual cortex (V1) to a two alternative forced choice (2-AFC) perceptual decision. We show that while both sparse and non-sparse linear decoding yield discrimination results quantitatively consistent with human psychophysics, sparse linear decoding is more efficient in terms of the number of selected informative dimension.

Highlights

How the brain represents information and how such a representation is utilized to form decisions and mediate behavior are fundamental questions in systems neuroscience, being addressed at many different scales from single-unit recordings to neuroimaging across the brain
Work in theoretical neuroscience has argued that a useful information processing strategy for the brain is to map sensory information into a sparse representation – the sparse coding hypothesis (Olshausen and Field, 1996a)
Perceptual decision making in a V1 model et al (2006), which was constructed by taking images from the internet, segmenting the car from the background, converting the image to grayscale, and resizing to be comparable as the face images

Summary

Introduction

How the brain represents information and how such a representation is utilized to form decisions and mediate behavior are fundamental questions in systems neuroscience, being addressed at many different scales from single-unit recordings to neuroimaging across the brain. Sparse coding has been viewed as an optimal strategy for minimizing redundancy and has been experimentally observed and theoretically justified for a number of sensory systems including visual (Baddeley, 1996; Dan et al, 1996; Baddeley et al, 1997; Vinje and Gallant, 2000, 2002), auditory (Hahnloser et al, 2002; Hromadka et al, 2008; Greene et al, 2009), olfactory (Perez-Orive et al, 2002; Szyszka et al, 2005; Rinberg et al, 2006), and motor (Brecht et al, 2004) systems It is considered efficient from a metabolic and energy perspective. Given such a sparse encoding strategy, how does the rest of the visual system utilize it for robust and efficient object recognition? How is the sparse representation exploited downstream to yield the behavior we observe?

Objectives

Methods

Results

Conclusion