Face motion and the 3rd dimension in keypoints in Motion-X++

[xuanlongORZ]

Hi, thank you for the great work, it really helps a lot !
I have two small questions on Motion-X++ dataset. When I go through the keypoints (which I mainly use), I find there is a 3rd dimension (so for one keypoint, it looks like [x, y,?]). I wonder if the 3rd dimension represents visibility. The second question is about the face motion in the subfolders (forgive me if it is a silly question) are they emotions with generation? or are the captured ones in the real world without many changes?

Looking forward to your feedback.