We're working on an assistive technology project with a built-in calibration system utilising The EyeTribe Calibration API. The calibration works fine, but we'd like to be able to show calibration result quality in terms of stars (out of 5) to users like what EyeTribeUI is able to.
As I see, this info is not present in the calibration result JSON. However, from my understanding, these results are probably derived from the average pixel or degree error. It's possible for us to experiment and come up with some classifications, but I guess that would not be very accurate. Would you mind share with us the accurate criteria/formula used to estimate the calibration quality ?
Highly appreciate your help.
Thanks a lot
)