Emotion Recognition System in Images Based On Fuzzy Neural Network and HMM

An emotion recognition system based on neuro-HMM was proposed to analyze the emotion contained in images. This system took an initial step in this direction by describing a set of proposed difficulty metrics based on cognitive principles. Both the emotion semanteme extraction and emotion model construction were considered in this system. They were respectively carried out by neural networks and HMM. According to the strong relationship between image notable lines and human dynamism sensation, the system used fuzzy neural network to establish the mapping and obtained the image emotion semanteme sequence. Then the duple hidden Markov model (HMM) was employed to simulate human emotion transition and finally confirmed different emotion models. The system also considered some outer influences to make the system rules be refined in realistic conditions. The experiment shows at least one emotion from an image can be recognized. The results illustrate the capability of the developing image recognition system