On the Use of Lateralization for Lightweight and Accurate Methodology for EEG Real Time Emotion Estimation Using Gaussian-Process Classifier

Abstract

Emotional estimation systems based on electroencephalography (EEG) signals are gaining special attention in recent years due to the possibilities they offer. The field of human-robot interactions (HRI) will benefit from a broadened understanding of brain emotional encoding and thus, improve the capabilities of robots to fully engage with the user’s emotional reactions. In this paper, a methodology for real-time emotion estimation aimed for its use in the field of HRI is proposed. The proposed methodology takes advantage of the lateralization produced in brain oscillations during emotional stimuli and the use of meaningful features related to intrinsic EEG patterns. In the validation procedure, both DEAP and SEED databases have been used. A mean performance of 88.34% was obtained using four categories of the valence-arousal space, and 97.1% using three discrete categories; both of them obtained with a Gaussian-Process classifier. This lightweight method could run on inexpensive, portable devices such as the openBCI system.