I would guess that the “foveated” region that they stream is larger than the human fovea, large enough to contain the saccades movement (with some good-enough probability).
Saccades afaik can jump to an arbitrary part of the eye which adds to the latency of finding the iris; basically the the software ends up having to look through the entire image to reacquire the iris whereas normally it’s doing it incrementally relative to the previous position.
Are you really sure overrendering the fovea region would really work?
Not sure, probably depends on the content too. When you read text, the eye definitely isn’t jumping “arbitrarily”, it’s clustered around what you’re focusing on. Might be different for a FPS game where you’re looking out for ambushes.
I’m not sure what you mean by “look through the entire image to reacquire the iris”? You’re talking about the image from the eye tracking camera?
> You’re talking about the image from the eye tracking camera?
Yes. A normal trick is to search just a bit outside the last known position to make eye tracking cheap computationally and to reduce latency in the common case.
