Emotion Recognition of EEG Underlying Favourite Music by Support Vector Machine
Published in IEEE International Conference on Orange Technology, 2013
This study aims to research the relationship between electroencephalography (EEG) at the prefrontal cortex (PFC) and emotion in the condition of different preference levels of music by applying a support vector machine (SVM). To achieve this, this study presents an EEG-based brain computer interface (BCI) music player, which can simultaneously analyse brain activities in real time and objectively provide therapists with physiological data for emotion detection in the experiment. The SVM result shows that more than 80% accuracy of elicited emotion based on 28 participants was analysed under the two factors of the frontal midline theta and alpha relation ratio. As such, it might suggest that significantly different stimuli are capable of enticing discernible EEG responses at frontal lobes, which is an indication of emotion and of providing an effective approach for application to multimedia with the abilities of EEG interpretation.
Kevin C. Tseng, Borshy Lin, and Chang-Mu Han, "Emotion Recognition of EEG Underlying Favourite Music by Support Vector Machine," IEEE International Conference on Orange Technology, 155-158, 2013.