Emotion recognition from EEG brain signals based on particle swarm optimization and genetic search

Abstract

The purpose of this study is to classify the emotions from human brain signals using electroencephalography (EEG). EEG signals were acquired while subject were watching the emotional stimuli. Subjects were asked to watch four types of emotional stimulus such as, happy, calm, sad and scared. The EEG signals were recorded using 14-channel brain headset. We preprocessed the EEG recorded data with manual artifact rejection and independent component analysis (ICA). The total numbers of 21 subjects were participated in this experiment. We performed emotion recognition which was based on two feature selection methods such as, particle swarm optimization (PSO) and genetic search (GS). Further, the selected features were processed using support vector machine (SVM). Emotion recognition accuracy had shown the possibility of classification of EEG brain activity.