The general objective of this postdoctoral project is to investigate how visual self-motion feedback, FOV constraints, and avatar characteristics shape embodiment, locomotor control, and social navigation in immersive populated virtual environments. Through close collaboration between VR engineers and computer scientists at Inria VirtUs and neuroscientists and researchers at the University of Waterloo’s Multisensory Brain and Cognition Lab, the project seeks to model and experimentally validate the multisensory computations underlying a coherent sense of self and presence in VR when navigating through crowded environments.