Comparison of ERP & Entropy based Emotion Classification using EEG signals

Emotions are necessary to understand the behavior of an individual. andrealistic decision-making. Since number of techniques can be used for emotion recognition such as voice, facial expression of an individual but these channels can be faked. In this paper, main emphasis is given toward the acquisition of EEG signal emotion evoking picturesprovided by International Affective Picture S decomposed into five different frequency bands namely delta (0 gamma (3264Hz) by using filtering technique. ERP potentials such as P100, N100 and the two latencies corresponding to these bio potentials emotion from preprocessed EEG signals. The training and testing is performed on both the attributes extracted from EEG for classification of emotions into two classes validation to classify the emotions along arousal axis. It related potential as compared to Entropy attribute. The maximum accuracy of 79.16% has been achieved along ERP attribute and 68.5% along Entropy attribute.