Towards emotion recognition of EEG brain signals using Hjorth parameters and SVM

There are several techniques of Psychophysiology data collation from human subjects. But, we focused in this paper to present the emotion detection of EEG brain signals using a classifier which is known as Support Vector Machine (SVM). The emotions were elicited in the subjects through the presentation of emotional stimuli. These stimuli were extracted from International Affective Picture System (IAPS) database. We used five different types of emotional stimuli in our experiment such as, happy, calm, neutral, sad and scared. The raw EEG data were preprocessed to remove artifacts and a number of features were selected as input to the classifiers. The results showed that it is difficult to train a classifier to be accurate over the high number of emotions (5 emotions) but SVM with proposed features were reasonably accurate over smaller emotion group (2 emotions) identifying the emotional states with an accuracy up to 70%.