Emotion Recognition from Physiological Signals

Nowadays keeping healthy has become one of the most important topics in our daily life. Keeping good mood is very helpful to one’s health. A lot of smart sensing systems have been designed and developed to detect human emotions. The physiological parameters obtained from the sensing system are then received and analyzed by computers. The physiological dataset collected by computers is then processed by 2 stages—feature extraction and classification. On the classification stage, datas are classified into different emotion groups, an automatic recognition of emotions is realized by this way. This paper aims to provide the feature extraction method and an algorithm using clustering techniques for classification. Cross validation is also used in our project. The k means clustering. to train a mood detection model can help us judge an individual’s emotion accurately. And this training model displays good results.