Resilient Face Expressional Recognitions Using Geometry and Behavioral Characteristics

Tremendous growth in the use of biometric technologies in high-security applications has motivated the requirement for highly dependable face recognition systems with expressional robustness. Current standards demonstrated that the accuracy of state-of-the-art algorithms is fairly high under constrained conditions, but degrades significantly for images exhibiting facial expression variations. The previous works try to achieve insensitivity to such variations leads to the motivation in classifying the physiological properties of the facial expressions and mapping it to the behavioral traits underlying the expression. The system developed in this paper describes and evaluates a resilient face expressional recognition system using a geometric structural representation of the various expressions like sad, angry, disguise, happy etc and mapping it to the behavioral traits stored in the form of either hidden or exposed property in the genes adopted in genetic algorithm model. The experimental evaluations are conducted with initial taking of 10 expression samples of each and every faces of 63 humans (21 men and 42 women). The geometric structure linking to the expressions are identified and store it the training samples template. The behavior pattern mapping of the expressions are associated with the genetic properties. The matching phenomenon is tested with the training template to evaluate the recognition and its rate of matching.