Asymmetric Facial Shape Based on Symmetry Assumption

It has been known that it is hard to capture the high-frequency components (shadows and specularities) during the modeling of illumination effects. In this paper, we propose a reflectance model to simulate the interaction of light and the facial surface under the assumption that face is strictly axial symmetry. This model works well not only in fitting the intensities of pixel but also in processing the DC component contained in the image. To compute a facial 3D shape, we first augment the input images to get a symmetric facial normal field, then propose a method to obtain a more accurate normal field, and finally compute an integrable shape using the field. Experimental results for face relighting, facial shape recovery demonstrate the effectiveness of our method.