Smile Emotional Identification for Negative Emotions Detection by Fuzzy Neural Network with Pixel Differences

In recent days, smile detection in face images is an interesting problem. The smiling face expresses many emotional expressions which is context dependent. That is the people can as well smile to convince other people. Thus, the felt and faked enjoyment facial expressions detection is a very important task. Earlier, many existing technique proposed to attain this goal, however, the interpretation of human smiles is often context dependent which reduce the detection accuracy. To solve this problem advanced prediction emotional contagion has been proposed to identify the emotion including negative emotions in the gray scale smiling face. Where the smile and non smile pixels are differentiated thorough the intensity values between pixels and considered as a feature value. Finally, using these features the classification task is achieved using Fuzzy Neural Network (FNN) classifier to classify smile and non-smile faces. Experimental results shows that FNN based emotion detection system has higher detection accuracy rate of 94% and computationally faster than the state-of-the-art method