Main thing I got seems to be that there's a more conventional multi-cam setup that can, seemingly, quickly generate a 3d model. I'm curious to see what happens

Main thing I got seems to be that there's a more conventional multi-cam setup that can, seemingly, quickly generate a 3d model. I'm curious to see what happens when someone does a hair flip. They show someone holding up a piece of paper, so it's got to be reactive, but the way they phrased the modeling made it sound like it's reusing a model/a retained mode, rather than rebuilding the world/a immediate mode rendering technology. How does the display tech here work? Anyone know?