Supporting live data sessions in multi-user virtual reality using CAVA360VR

A dense composite 2D video clip showing the different screens and audio chat of three analysts using CAVA360VR (Collaborate, Annotate, Visualise, Analysis 360 video in Virtual Reality) to hold a live virtual data session at a national research conference in 2019. The three co-analysts (M, J and P) in VR were collaborating even though they were in three different physical locations across two countries. The video data that they analysed together involved multicam recordings of six participants in a two-team competition to solve a Lego puzzle when only one member in each group could view the hidden Lego construction in order to instruct the others in their team how to rebuild it from a pile of blocks. The video clip shows the viewport chosen by each of the three analysts. There are three possible viewports that can be selected: a) what they each see in their VR HMD (head-mounted display), b) an over-the-shoulder view (with an outline showing their avatar), and c) a static third-person view of the three analysts. A fourth screen shows the view of an observer in non-VR mode.