FIX: feature-based image transformations for face animation

We propose a simple yet effective 2D image transformation method for face animation. Instead of using complicated 3D models or a large database of 2D images, a set of transformations are learned to create different visual effects in a given image, including talking, changing facial expressions, and head movement. This approach enables creation of realistic images with minimum input data and efficient computation. Scaling and mapping these transformations into different head orientations is discussed. The ShowFace system, a modular streaming framework for face animation, is also introduced which uses the proposed technique.