{"localization":"Easy up until you start doing the part below the eyes. It requires temporaries and guess-and-check or 2-d logic. The mouth is ambiguous (2 possibilities i think) and you just have to guess based on the image.","ok":true}