Joint Confidence Always 1 in Pose Annotation

[duremovich]

Hi all!

First of all, this project is awesome. I'm a projections designer for dance and am super excited for the possibilities this framework provides.

My question is about the JSON joint data from pose annotation. For some reason my confidence for every joint is always 1. Unwaveringly so. Even if the camera can't see the joint at all. All the x, y, z coordinate data is working accurately.

I'm curious if I missed a setting somewhere or if this piece of a data is just a placeholder for future functionality that isn't implemented.

Thanks!

Dylan