The goal is to improve some of the current state-of-the-art in 2D face image processing, specifically developing improved algorithms for obtaining the positions of faces and parts of it (eyes, mouth, nose) in noisy images where the face can be partilly occluded.