About MPII dataset

[CYH4157]

I think your method is very powerful after reading your paper.
I would like to ask a question. I would like to implement your end-to-end method, but I don't seem to see any spii dataset using end-to-end training in your repo. I may have missed it, where would it be in this repo?

[likenneth]

COCO pose estimation is a multi-person pose estimation dataset; MPII is a person pose estimation dataset. In MPII, there is no need to detect a person as a first step. Therefore, only the second Transformer is needed, and no end-to-end model is needed.