How did you compare speed with pix2pixHD in the paper? #11
Description
Because pix2pixHD need to train global and local model, how did you train the pix2pixHD model when you comparing speed with your method?
Did you train pix2pixHD global model in 256*256?