Face Alive Icons

Facial expression is one of the primary communication means of the human. However, realistic facial expression images are not used in popular communication tools on portable devices because of the difficulties in: 1) Acquisition; 2) Transference; 3) Display. In this paper, we propose a system tackling these problems to synthesize facial expression images from photographs for the devices with limited processing power, network bandwidth and display area, which is referred as “LLL” environment. The facial images are reduced to small-sized face alive icons (FAI). Expressions are decomposed into the expression-unrelated facial features and the expression-related expressional features. The common features are captured and reused across expressions by the discrete model built through the statistical analysis on the training dataset. Semantic synthesis rules are also constructed which reveal the inner relations of expressions. Verified by an experimental prototype system, the approach can produce acceptable facial expressional images utilizing much less computing, network and storage resource than the traditional approaches. 1