On the same subject, I prefer the concurrently published "Image Analogies" paper ( http://www.mrl.nyu.edu/projects/image-analogies/ http://www.mrl.nyu.edu/proje

On the same subject, I prefer the concurrently published "Image Analogies" paper ( http://www.mrl.nyu.edu/projects/image-analogies/ http://www.mrl.nyu.edu/projects/image-analogies/ ) because their described technique, while slow, just makes so much sense.