Image Morphing

Statement

A morph is a simultaneous warp of the image shape and a cross-dissolve of the image colors. The cross-dissolve is the easy part; controlling and doing the warp is the hard part. The warp is controlled by defining a correspondence between the two pictures. The correspondence should map eyes to eyes, mouth to mouth, chin to chin, ears to ears, etc., to get the smoothest transformations possible.

Defining correspondence

Using a consistent definition of correspondence between two images, we can define a triangulation for morphing.
Generating correspondence between images with keypoints and Delaunay triangulation.

Computing the Mid-Way Face

Computing the mid-face takes two images with corresponding keypoints and a triangulation. The images are morphed into average geommetry and cross-dissolved to get the final image.

Bells-Whistles

Morphing George to Morgan with a sequence of 45 frames.

Mean-Face from a Population

Using the IMM dataset, we compute the mean for a subset of male-population images.
Some faces projected to mean geometry.
The mean population face.
Projecting George's face to mean geometry of male-faces.
Projecting mean face to George's geometry.

Extrapolation

With the mean population face, we can extrapolate the mean face to our target frame.
Caricatures of George from mean population face.

Bells-Whistles

Using mean face for African-American ethnicity.