8/09/2016 - 11:00 Mr Jonathan Vacher (Université Paris-Dauphine) Salle 106 - Batiment IMAG
Vision is widely studied along the visual pathway from the retina to the higher visual areas. Most generally, the goal is to relate a stimulus to the data collected during an experiment. We introduce a class of relevant visual stimuli and some application in psychophysic, neuroimaging and neurophysiology. The stimuli consist in space-time Gaussian random fields with a parametric covariance and the experiments aim at testing some aspect of this covariance in order to exhibit a possible influence in the observed data. These data allow many approaches from Bayesian inference to supervised learning, we will show how these methods can be useful to interpret this various kind of data.