An Emotion Classification Method for Individuals Using EEG and Heart Rate Data and Deep Learning

Abstract

In recent years, techniques to classify emotions by engineering have been studied and applied to various fields. Among them, the emotion classification using biological information is classified by using the EEG of the central nervous system and the heart rate of the autonomic nervous system. However, this emotion classification has problems such as not considering individual differences. In this study, we aimed to classify emotions by considering individual differences by learning with deep learning using EEG and heart rate as input data. In the proposal, we tried to classify four emotions by devising a method of acquiring subjective emotion data, which is the correct answer data. As a result, we were able to classify emotions. Furthermore, analysis of the input data suggests that the heart rate may be an important feature in emotion classification, suggesting the need to use both EEG and heart rate for emotion classification.