Facebook is looking at solving one major problem with VR headsets, they isolate the user from their environment and, in doing so, limit VR use and acceptance.
Eliminating this isolation is a key motivation for the development of video pass-through VR, wherein the VR headset user sees
a reproduction of their external environment and the individuals within it. Yet, a crucial gap remains: External viewers cannot hold a natural conversation with a VR headset user, whose upper face and eyes remain occluded.
the full paper by Nathan Matsuda, Brian Wheelwright, Joel Hegland, Douglas Lanman can be found HERE.
Facebook is looking to develop reverse pass-through VR, wherein a three-dimensional view of the wearer’s eyes is presented to multiple outside viewers in a perspective-correct manner, with a prototype headset containing a world-facing light field display. This approach, in conjunction with existing video (forward) pass-through technology, enables more seamless interactions between people with and without headsets in social or professional contexts.
Reverse pass-through VR ties together research in social telepresence and copresence, autostereoscopic displays, and facial capture to enable natural eye contact and other important non-verbal cues in a wider range of interaction scenarios.
Prior findings lead to the belief that reverse pass-through VR systems must use an autostereoscopic display. Many autostereoscopic displays have been proposed, but Facebook is advocating that light field displays based on microlens array (MLA) technology are most compatible with the current research and industry trends toward thinner headset form factors. Recent advances in polarisation-based optical folding or “pancake” viewing optics have established a path toward significantly reduced headset volumes.