Thursday 9 October 2014, 4.00PM to 5.00pm
Speaker(s): David Watson
A hypothesis- and data-driven investigation into the role of low-level visual cues in scene processing
Human ventral visual cortex has been noted to contain regions showing selectivity for scenes, including the Parahippocampal Place Area (PPA), Retrosplenial Complex (RSC), and Occipital Place Area (OPA)
Here we propose two fMRI experiments investigating the role of low-level visual properties in the representation of scenes within such regions. In the first case we propose a hypothesis driven experiment investigating how neural representations of scenes are affected by disruptions to the local statistics of the images. In the second case we aim to measure the neural response to “clusters” of scenes defined using an entirely data-driven approach based upon the relative similarity of their image properties. In both cases multivariate pattern analysis (MVPA) will be used to investigate how such manipulations influence the representation of scenes in the human brain.