Emotion Recognition with Wavelet Transforms from EEG Signals

Abstract

Nowadays, emotion recognition studies have gained importance in order to make the interaction between human and computer easier and more natural or similar to human-human interaction. This study proposes a new emotion recognition system based on wavelet transform of electroencephalogram (EEG) signals sensitive to music videos. Wavelet coefficients are obtained by applying Wavelet Transform to EEG signals. To reduce the size of these coefficients, the feature vector is created by applying Principal Component Analysis and Information Gain techniques. The resulting reduced property space is tested by various classification algorithms. The proposed method is compared with other studies using the same data set in the literature and the results are examined in terms of performance. © 2019 IEEE.