Yoonho Shin

Difference between results from inference and the paper

25

First, thanks for your great work. I trained the model using script 'python train.py --gpu 0-1 --backbone LPSKI' with Human3.6M and MPII datasets. The protocol is 1 and train epoch...