Key points are not available for this paper at this time.
The recent development of new video capture systems has led to the adoption of volumetric video technologies to replace 2D video in use cases such as videoconference, where this enhancement promises to solve videoconference fatigue. In particular, volumetric capture allows the content to be viewed from different points of view, enabling more natural interaction during the videoconference. One of the solutions proposed for this scenario is Free Viewpoint Video (FVV). It makes use of a set of calibrated cameras that allows the use of real life information to generate a synthetic view from any arbitrary point in space. Although there are real-time capture developments of FVV systems, they make use of 2D displays and joysticks to control the point of view. In our opinion, this undermines the possibilities of volumetric video for the videoconferencing use case. Building on a previously developed FVV system, we present a novel untethered HMD-based immersive visualization system that enables point of view control with the user's natural position and visualization of live volumetric content in a 3D environment. Synthetic views are generated in real-time by the FVV system, and streamed with low latency protocols to a Meta quest 3 HMD using a WebRTC-based server. This work discusses the architecture of the end-to-end system and describes the bitrate, framerate and latency values at which the system works.
Usón et al. (Thu,) studied this question.