Séminaire du GREMLIN - Victor Rosi

Hearing one’s own recorded voice is often described as an uncanny or unsettling experience. This phenomenon is typically explained by a mismatch between externally heard recordings and the multisensory (i.e., auditory, vibrotactile, bone-conducted) perception of the self-voice during speech production. Yet, the recorded self-voice remains a highly familiar and self-relevant stimulus, which confers specific perceptual advantages. In this talk, I will present a series of recent behavioural studies investigating the unique perceptual and social status of the self-voice. Specifically, I will present how listeners process and evaluate their own recorded voice compared to other voices. Along the way, I will also examine whether this self-related advantage extends to artificially generated voices, exploring the extent to which modern voice cloning technologies - now increasingly present in everyday life - preserve or disrupt the specificities of the self-voice. Overall, this work contribute to our understanding on how we perceive self-relevant information in an era where recorded voices are ubiquitous and AI-generated replicas increasingly blur the boundary between real and synthetic identities.

Le séminaire aura lieu en hybride, et pourra être suivi à partir du lien suivant : https://meet.google.com/iru-wmri-oey