Emotion recognition based on correlation between left and right frontal EEG assymetry

The significant role of emotion recognition research has increased in last few years of human daily life. In this paper, we base on electroencephalogram (EEG) of the brain to recognize the internal emotion of participants. We use arousal and valence emotion elicitation process to calculate multidimensional direct information (MDI) between right and left hemisphere. The main contribution is recognizing internal emotion of people based on reading the signals from frontal asymmetry of brain so that four channels are used to represent frontal asymmetry of brain F3, F7, F4 and F8. The emotion elicitation process is focused on frontal EEG asymmetry base on using standard dataset. We test four basic emotions happy, sad, anger and relax. These emotions represent high arousal high valence, low arousal low valance, High arousal low valence and low arousal high valence in arousal and valance model. Statistical-based features from EEG signals of data inputs are extracted and used to calculate multidimensional direct information in order to find the correlation between left and right hemisphere to each emotion.