I combined my face with my brother's face. First I find the mean points between the shapes of our faces. Then I compute a delaunay triangle on those mean points. For each triangle, I find the affine transformation from the original image's triangle to the average triangle and transform all points in the triangle using that transformation. Thus, I morph my face and my brother's face to the average shape and then cross dissolve by taking the average of those two morphed images to get the midway face.
Here is my morph sequence where I morph both images to an intermediate image based on a weighted mean of my face vs. my brother's face using fractions from 0 to 1 and combine all the intermediate images to create my morph sequence.
I took the average shape of women from the Danes data set with neutral expressions using their feature points. Here are some of the original images and then those images morphed to the average shape, morphed using the same triangulation method.
Here is the average face of neutral expression women, by cross dissolving all their faces morphed to the average shape.
Here is my face morphed to the average shape and the average face morphed to my shape. I think the warp looks especially weird because I'm smiling whereas the other women aren't.
Here is my face by extrapolating from the mean with an alpha of 1 and then 1.5, which I do my adding the difference between my face and the mean multiplied by the alpha.
I made a morph sequence with my whole family. I morphed the images pair by pair and combined all the resulting images.
I morphed my friend's face to look more female by using the averages from the Danes females. I first morphed his face shape to the average shape (using triangulation/affine transformation), combined his appearance with the average appearance (using cross dissolve), and then did both
