Study of Emotional Variability Using Photoplethysmogram Signal

Abstract

This study investigates the feasibility of photoplethysmogram (PPG) signals in recognizing variability in human’s funny, fear and sad emotions. Undoubtedly, Easy Pulse data acquisition device which is used to perceive the PPG signals have superior criterions which are small in size, low power consumption as well as low in cost. Thus, this study will prove the robustness and reliability of PPG signals as an emotion recognition mechanism. A total of ten subjects were chosen randomly which ranged from twenty-one to twenty-four years old. A total of five male and five female students were given three different videos to stimulate different emotions during the given time. Easy Pulse sensor, which has the ability in filtering the unwanted signals has made the study easier. Discriminative features are then extracted from the PPG morphology. PPI, maximum amplitude, as well as the Cardioid pattern of the signals. Finally, four methods of classification have been used to identify the variability in emotions. PPI, maximum amplitude, area and maximum radius of the Cardioid loop were used as the classifiers. These methods have clearly shown great results in differentiating between funny, fear and sad emotions. It was discovered that every human has different rate of sensitivity to fear and sad. Some have the tendency to be very sensitive to fear and some to sad. The experimental results demonstrated that the physiological signals such as PPG have great potentials where the system provides high classification performance.