Throughout this project, I labeled 43 points of correspondences by hand for each non-annotated image. I overlayed the midway points and triangulation to create a visualization of the correspondences on each image.
Face A points
Face B points
Computing the "Mid-way Face"
For this sharp image, I blurred it with a Gaussian filter first, and then sharpen the blurred image. The result is somewhat sharper than the blurred image on bright edges, but it is not close to the original image. We cannot recover the original resolution using this method.
Face A
Mid-way Face
Face B
The Morph Sequence
Morph Sequence
The "Mean face" of a population
I used the Danes collection and computed the average of all faces to get the mean face.
Mean face of Danes collection
Morphing each of the faces into the average shape
Original (01-1m)
Morphed into average shape
Original (06-1m)
Morphed into average shape
Original (06-1m)
Morphed into average shape
Morphing my face into the mean face
My face warped into the average geometry
Average face warped into my geometry
Caricature
I used alpha = -0.5 to extrapolate my face away from the mean
Bells and Whistles
I participated in the group morphing video with other students! (https://youtu.be/O2p-U615Rzs)
In addition, I made a face morphing video on the theme of growing up, which records pictures of myself from different times and morphin the faces through time. (https://youtu.be/okPI3En-ARw)
I also played around with changing the gender of my face.
