Gender in Face-Voice Integration (Face-Voice)

Bayesian models of perception represent a promising approach to describe information processing by the brain. Predictive coding hypothesizes a process in which top-down expectations are continuously compared across multiple hierarchical levels with bottom-up sensory inputs and the differences or error signals are propagated in a bottom-up direction. The investigators hypothesize that strong expectations, are best investigated in expert processes such as face and voice recognition in humans. Individuals in complex social systems need to extract socially relevant information in a fast and efficient manner; hence, the majority of humans constitute face, voice and gender experts. Nevertheless, linking such combined abilities to brain activity with regards to the predictive coding hypothesis has not been attempted.Our results suggest asymmetric contributions of visual and auditory signals to the gender classification task. This sensitive psychophysical procedure is implemented in a decoding approach using fMRI and multi-voxel pattern analysis (MVPA). The investigators plan to test whether cortical areas implicated in processing auditory and visual gender signals show similar asymmetries.