Emotion detection using psycho-physiological signal processing

The purpose of this paper is to assess the value of physiological responses as specific indicators of emotion. Our study consisted of 4 stages: emotion induction, signal recording, signal processing (feature extraction) and pattern classification (detection). The emotion induction protocol was based on the presentation of emotionally-evocative images from IAPS (The international Affective Picture System) to the subject. The signal used to analyze physiological response is EDA (electrodermal activity). We were able to differentiate between 2 classes of emotion: pleasant and unpleasant.