The first step in the procedure is to select points and find a triangulation of the images. For this step I manually input the corresponding points and I used the Delauney traingulation of the points.
Using the same algorithm, I found the average of the points in the "Danes" face data set and warped each of the images to the average. Then I averaged each warped image to create the "average face."
I calculated warps between me and my roommates, however there is a lot of ghosting. This is because we have very different hairstyles and I did not select points on our hair and shoulders.