Another potential thing as well: Before watching the video it was instinct for me to try to Swipe to move on. Kinda like a eBook. Do you think there is a way to incorporate that to the tracking?
I actually had point finger left and right gesture recognition, as well as just have a hand at the edge of the screen left or right to move, but caused unexpected movement to people so removed it last minute. But swiping is a good idea. Doable :)