There are already some open issues regarding the paired image to image translation task. I am also interested in this. The authors present also comparisons with improved pix2pixHD variants such as pix2pixHD++ w/ SPADE and a compact version.
Are these models available? I could imagine that an improved pix2pixHD version would be interesting to many.
Alternatively, what is the current state of the art model for supervised image to image translation? Most research seems to focus on unsupervised, unpaired translation tasks, or do I see this wrong?
There are already some open issues regarding the paired image to image translation task. I am also interested in this. The authors present also comparisons with improved pix2pixHD variants such as pix2pixHD++ w/ SPADE and a compact version.
Are these models available? I could imagine that an improved pix2pixHD version would be interesting to many.
Alternatively, what is the current state of the art model for supervised image to image translation? Most research seems to focus on unsupervised, unpaired translation tasks, or do I see this wrong?
Thank you.