Rule based Web Interference Model for Volumetric Facial Feature Extraction in Facial Emotion Detection using Neural Networks Krishna E. M. Ashok*, Dr. Dinakaran K.** *Research Scholar, Department of Computer Science, Karpagam University, Coimbatore, Tamil Nadu, India **Principal, PMR Engineering College, Chennai, Tamil Nadu, India Online published on 15 September, 2016. Abstract Facial emotion detection is an upcoming research which has more importance in applications like person identification and authentication systems. Emotion detection has been done used different features like shape, color, farat and facial components which work only offline. We propose a new rule based inference model for facial emotion detection using volumetric estimation of facial components. The proposed method captures the human face through the web cam and handed over to the web server. There the proposed method removes the skin regions using intensity approximation method and the facial components like eye, nose, mouth features are extracted using template matching techniques. The extracted features are passed through neural network where each neuron computes the volumetric estimation with the given input point. The input and output value of each neuron function is passed through different levels and finally the neuron produces an output as emotion using the rule sets available. The proposed method has more advantage than offline methods by using more number of templates and has produced accurate results. Top Keywords Facial Detection, Emotion Detection, Rule Based Model, Web Inference, Neural Network. Top |