Facial Expression Detection using Pattern Analysis and Machine Intelligence

Emotion is a mental state that happens spontaneously rather than through conscious attempt and it is accompanied by physical changes. Human computer interaction can be automated by using emotions and the accompanied physiological changes. The computers not only respond to human commands, they also try to mimic the human behavior. Automatic recognition of human facial expression and emotion takes the computer one step near the above said task. Human emotion recognition is an interesting and challenging problem. In order to translate facial expressions to emotions, an emotion assessment technique are required. Emotion assessment is difficult task even users are not always able to express their emotion with words all the time and the self-reporting emotions have a high probability of false emotions. In this work, initially the human face will be detected. Then the facial features will be extracted and classified into different expressions. The features extracted from the image will be quantified and will be used as training set for the pattern recognizing neural network. The trained neural network in future will classify the images according to the emotions expressed by the person.