Facial expressions analysis based on cooperative neuro-computing interactions

Abstract

Facial expressions analysis is an important step in human-computer interaction and intelligence computing. Due to the complexity and uncertainty of real-time facial expressions, the performance of the existing algorithms is not satisfactory. In this paper, a novel approach is proposed to help to enhance the significance of analysis optimum. Based on the person-independent approach and the cooperative neuro-computing, multi-model interactions for facial expressions cluster structures are applied to improve the capacity of selection, distribution, and evaluation of the cluster centers. The resultant model is potentially capable of constructing high-quality clusters and achieving high efficiency of the convergence. It is suggested that the model with cooperative neuro-computing interaction has the characteristics to construct the cluster distribution rapidly, and can perform real-time analysis efficiently and accurately. © Springer-Verlag 2013.