I just had a look. Is the user supposed to create the depth map? I do this for my stereograms. But building a depth image from a photo is much more difficult.

I tried several photos and each one was rotated 90 degrees.