joints_2d are not exact

[andreaskoelsch]

Hi,
I just downloaded the dataset and ran some first inspections. On the very first frame I checked, I noticed that the joints_2d are a bit off. The sequence I am referring to is FPS-30/2020-05-20-21-13-13/

It looks to me like the projection matrix is wrong, so I computed the joints_2d from the 3d positions myself, but with the same result.
Here's a visualization of the keypoints.

Is there a way to find out which frames are annotated correctly and which ones are not? Or even better, can the annotations be fixed?

[andreaskoelsch]

I randomly looked at some other samples to find out if this error was just an outlier that I happened to come across at the first sample I looked at. But unfortunately, I found these bugs in many other samples. Here are some other quite severe one, that are really beyond "measuring tolerance".