Emotion Detection Based on EEG Signal

One of the foremost critical element between interacting individuals is emotion.
Today, it is vital for the computers to recognize the emotion of user that interacts with
the computer in Human Computer Interaction (HCI) frameworks to make the devices
more successful and comprehensive for everyone. In our body, signal from our brain,
also known as electroencephalogram (EEG) signal is the main source that generates
emotion. Lately, emotion detection through EEG signals had pulled in numerous
researchers and numerous algorithm were discovered. Various kinds of features
extraction were investigated and different types of classification method were explored.
However, emotion detection based on EEG signal is considered as a challenging
research topic due to the non-stationary behaviour of the signal. Thus, this project aimed
to study the emotion detection through EEG signal and proposed the right algorithm to
process the signal. In this research, two class of emotion which are happy and sad are
detected through EEG signal. Wavelet transform scalogram is used as feature extraction
method. After that, the signal go through Convolutional Neural Network (CNN) for
classification.