Item

Abstract

Studies on visual processing often rely on highly artificial stimuli such as bars or gratings. As a result, little is known about the neuronal coding of those stimuli that we have to process on a daily basis. Addressing this issue, we analyze population responses (local field potentials, fMRI-BOLD, and spiking activity) in the cat visual cortex to natural scenes and noise stimuli with modified spectral statistics. We find that activity patterns elicited by the classical lab stimuli are qualitatively and quantitatively different from those elicited by natural stimuli. Investigating the variability of neuronal activity, we find that the (complex) temporal profile of natural scenes imprints itself on the local field potential. Exploiting this, we demonstrate how multiple properties of the stimulus are encoded simultaneously in the population response. Together, our results demonstrate that using oversimplified stimuli leads to similar oversimplified response patterns, which miss much of the structure that neuronal responses have during every day vision.