Charly Lamothe (PhD Student)

PhD Project :

In the context of a comparative study of voice perception in primates (ERC project COVOPRIM led by Prof. Pascal Belin), we will study in this PhD the contribution of machine learning and deep learning techniques for the exploration of the “vocal brain”. These techniques will be used to characterize the representation of voice information at different nodes of the network of voice-sensitive cerebral areas, with the aim of identifying higher-level features of the vocal signal that map quasi-linearly to brain activity. An important milestone towards this goal will be the reconstruction of voices from cerebral activity measured by functional magnetic resonance imaging (fMRI) and/or magnetoencephalography (MEG) during stimulation with these voices. Moreover, fMRI measures of sound-induced activity will be obtained in both humans and non-humans primates (macaques and marmosets), which will allow us to compare voice representations between species based on the higher-level features identified. Such approaches have proven themselves successful in multiple studies of the characterization of visual information processing in the brain, and increasingly start to be implemented in the auditory domain.