Automatic estimation of 2D facial muscle parameter using neural network

Muscle based face image synthesis in one of the most realistic approach to realize a life-like agent in a computer. The facial muscle model is composed of facial tissue elements and muscles. In this model, the forces affecting the facial tissue element are calculated by contraction of each muscle strength, so the combination of each muscle parameter decide a specific facial expression. Now each muscle parameter is decided on trial and error procedure comparing the sample photograph and the generated image using our Muscle-Editor to generate a specific face image. We propose the strategy of automatic estimation of facial muscle parameters from 2D marker movements using a neural network. This is also a 3D motion estimation from a 2D point or flow information in the captured image under the restriction of a physics based face model.