Characterization of Visual Object Representations in Rat Primary Visual Cortex

Sebastiano Vascon,Ylenia Parin,Eis Annavini,Mattia D'Andola,Davide Zoccolan,Marcello Pelillo
DOI: https://doi.org/10.48550/arXiv.1810.01193
2018-10-02
Abstract:For most animal species, quick and reliable identification of visual objects is critical for survival. This applies also to rodents, which, in recent years, have become increasingly popular models of visual functions. For this reason in this work we analyzed how various properties of visual objects are represented in rat primary visual cortex (V1). The analysis has been carried out through supervised (classification) and unsupervised (clustering) learning methods. We assessed quantitatively the discrimination capabilities of V1 neurons by demonstrating how photometric properties (luminosity and object position in the scene) can be derived directly from the neuronal responses.
Computer Vision and Pattern Recognition
What problem does this paper attempt to address?