Overview
For this project, we explore the idea of face morphining and creating animations by a simultaneuous warp of the image shape and a cross-dissolve of the image colors.
Defining Correspondences
For this part, we manually select corressponding points in each of the two images (Clooney's and my image) and use Delaunay Triangulation to find the most optimal triangles.
|
|
|
|
|
|
Computing the "Midway" Face
To compute the midway face, we set the warping fraction = 0.5 and the cross dissolve fraction = 0.5. So this takes the avg shape of the two faces and the average color in the two corressponding sets of triangles.
|
Morph Sequence
To generate a full transition we simply just use tweak the warping fraction and the cross dissolve fraction. I generated 45 frames. Both the warp fraction and the cross dissolve fraction are set = 1 in the frst frame, so we only see my picture and then are gradually increased to reach 0 in the last frame, to show Clooney's original picture.
|
The Mean Face of a Population
I used the Danes computer scientists dataset to calculate the mean face of a ppopulation. The mean face of a population is helpful to understand the biases in the population. As we see below, the avg danish computer scientist is a white male.
|
|
|
|
|
|
|
|
I also warped my own face into the average Danish face shape and morphed the average Danish face into my face shape.
|
|
|
|
|
Caricatures
Here we extrapolate using the points using this formula: pts1 + alpha*(pts1-pts2) and get the following caricatures.
|
|
|
|
Bells and Whistles
The first B&W I tried was changing my ethinicity. I got the image of the average Han Chinese Male from the Internet and tried to morph that into my face. Here are the results.
|
|
|
|
|
The next Bells and Whistles I implemented was participating in this class video which can be seen below. Here is a link to the video.
The next Bells and Whistle I implemented was making a video of me growing up. Here's a cute gif
Credit for the webstie template goes to Charlie Huang.