Eye synthesis using the eye curve model

Eyes are a critical part in exhibiting facial expressions. Because of the appearance diversity of eyes due to motion, it is difficult to synthesize eye with a particular facial expression. Traditional methods have failed to adequately catch motion-related appearance changes. In order to generate a photorealistic expression eye, we propose a two-step method. First, we propose a curve-based model to represent eyes. The model uses one circle and four skewed elliptical arcs to represent the shape of eyes, and divides the entire eye region into six sub-regions that correspond to different anatomical components of eyes. Then we propose a structure-based-similarity (SBS) framework to synthesize expression eyes using the eye curve model. The main contributions of this paper are: first of all, the proposed eye curve model can represent the diversity of eyes due to motion and structure, which is better than some traditional models. Second, the SBS framework is flexible. In multiple samples and multiple targets situation, the framework can synthesize eyes which exhibit same expression with different personal style. In single sample and single target situation, the framework can clone this expression successful. Experimental results show that synthesized eyes are realistic and expressive.