Face detection and facial feature extraction using color, shape and symmetry-based cost functions

We describe an algorithm for detecting human faces and facial features, such as the location of the eyes, nose, and mouth. First, a supervised pixel-based color classiier is employed to mark all pixels that are within a prespeciied distance of \skin color," which is computed from a training set of skin patches. This color-classiication map is then smoothed by Gibbs random eld model-based lters to deene skin regions. An ellipse model is t to each disjoint skin region. Finally, we introduce symmetry-based cost functions to search the center of the eyes, tip of nose, and center of mouth within ellipses whose aspect ratio is similar to that of a face. Face detection facial feature detection image segmentation shape classiica-tion Gibbs random elds