Thank you for making this repository available.
While reading the code, I noticed that the original annotations are provided as
[frame number, object type, object ID, behaviour index, occluded, x1, x2, y1, y2],
but the processed data contains 3D coordinates such as
[7593.49, 25702.70, -11903.41].
I would appreciate it if you could briefly explain how these 3D coordinates are obtained from the original 2D annotations.
If there is any documentation or relevant part of the code that describes this process, I would be grateful if you could point me to it.
Thank you very much for your time and your work.