In your paper, the edge2face experiment with CelebA-HQ was done with facial contour maps.
How did you make facial contour map?
It seems to me that this is created by concatenating landmarks of a face.
If it is true, what landmark detector did you use?
In your paper, the edge2face experiment with CelebA-HQ was done with facial contour maps. How did you make facial contour map? It seems to me that this is created by concatenating landmarks of a face. If it is true, what landmark detector did you use?