Advanced Automatic Human Emotion classification system with Lip Features Using M3 Filtering and Convolution Neural Network (CNN)

Human emotions play a major role in any individual life. There are many kinds of emotions like fear, happy, sad, hungry, surprise, pain and many more. Sometimes it is not possible to share few emotions happens. Some emotions may lead to suicidal tendency for few individuals. It is very important to trace out human emotions by others. Human emotions can be identified with facial expressions and sometimes few emotions can be found with lip gestures also. Here an automated system which identifies the human emotions with lip gestures is proposed. Proposed system is trained with Convolution neural network (CNN) with the features extracted from lip gestures. Proposed system can help in many aspects in present day situations. Accuracy for proposed system achieved is 95%.