Comparison of Emotion Recognition Methods from Bio-Potential Signals

This paper proposes an emotion recognition system from multi-modal bio-potential signals. For emotion recognition, two types of classifier: neural network (NN) and support vector machine (SVM) are designed and investigated. Using gathered data under psychological emotion stimulation experiments, the classifiers are trained and tested. In computational experiments of recognizing two emotions: pleasure and displeasure, recognition rates of 62.3% with the NN classifier and 59.7% with the SVM classifier are achieved. The experimental result shows that using multi-modal bio-potential signals is feasible and that NN is comparably more suited for emotion recognition tasks.