Neural representation of geometry and surface properties in object and scene perception

Matthew Lowe, Jason Rajsic, Jason Gallivan, Susanne Ferber, Jonathan Cant

Research output: Contribution to journalArticlepeer-review

13 Citations (Scopus)
8 Downloads (Pure)

Abstract

Multiple cortical regions are crucial for perceiving the visual world, yet the processes shaping representations in these regions are unclear. To address this issue, we must elucidate how perceptual features shape representations of the environment. Here, we explore how the weighting of different visual features affects neural representations of objects and scenes, focusing on the scene-selective parahippocampal place area (PPA), but additionally including the retrosplenial complex (RSC), occipital place area (OPA), lateral occipital (LO) area, fusiform face area (FFA) and occipital face area (OFA). Across three experiments, we examined functional magnetic resonance imaging (fMRI) activity while human observers viewed scenes and objects that varied in geometry (shape/layout) and surface properties (texture/material). Interestingly, we found equal sensitivity in the PPA for these properties within a scene, revealing that spatial-selectivity alone does not drive activation within this cortical region. We also observed sensitivity to object texture in PPA, but not to the same degree as scene texture, and representations in PPA varied when objects were placed within scenes. We conclude that PPA may process surface properties in a domain-specific manner, and that the processing of scene texture and geometry is equally-weighted in PPA and may be mediated by similar underlying neuronal mechanisms.
Original languageEnglish
Pages (from-to)586-597
JournalNeuroImage
Volume157
Early online date21 Jun 2017
DOIs
Publication statusPublished - 15 Aug 2017
Externally publishedYes

Fingerprint

Dive into the research topics of 'Neural representation of geometry and surface properties in object and scene perception'. Together they form a unique fingerprint.

Cite this