In my project, I developed a relatively simple facial model which can be rendered in a number of methods; at first, I used a simple two-dimensional ``cartoon'' model which was rendered using PostScript [7] code. Later, I took code from Andrew Marriott's implementation of Parke's model [3], adapting it to my model.
From the start, my facial model was designed as a generic, high-level model, which may be rendered in a number of ways. I have assumed that, in the applications of computer graphics and imaging, the face is firstly an expressive device, whose purpose is to transmit information in a language of sorts, and only secondarily a physical object. Working from this assumption, I developed my model as a top-down perceptual model.