IVOR · Neuronal substrates of invariant visual object recognition in rats

One of the key issues in the field of system neuroscience is to understand how the mammalian visual system is able to recognize visual objects in spite of tremendous variation in their appearance (e.g., lighting and viewpoint). So far, this issue has been mainly investigated in non-human primates, while simpler model systems (such as rodents) have been largely overlooked. However, recent work of Dr. Zoccolan and colleagues has shown that rats are fully capable of invariant recognition, rising the question of what neuronal machinery endows them with a visual processing ability that heretofore has only been ascribed to a select few species. Our project aims at addressing this question, by measuring the degree of overlapping/tangling of the neuronal representations of distinct visual objects across rat’s known (i.e., occipital), as well as putative (i.e., temporal) visual cortical areas. Our approach will be based on a combination of psychophysics, multi-units neuronal recordings, early genes dual-activity mapping, and machine learning tools. Our hope is to understand where and how, in the rat cortex, the gradual untangling of object representations happens, and to shed new light on a fundamental process that, in spite of several decades of monkey studies, remains elusive and poorly understood.