An Approach to Analysis of Physiological Responses to Stimulus: From Electrodermal Activity to Combined Physiological Responses

Abstract

Recent advances in wearable devices capable of measuring physiological signals such as Electrodermal Activity (EDA) support affective computing and related applications. We present an algorithm that uses EDA signals to detect highlights of a stimulus. To test the accuracy of our method, two different mediums, a scene from a movie, as our ground truth, and a scene from a video game, for testing the algorithm, were selected. We conducted a user study with 20 participants, analyzed the differences between mediums and validated the accuracy of our method for detecting the highlights of the stimulus using only EDA signals. Our approach uses commonalities among users based on their phasic responses for detecting highlights. The result of the study shows a F1 score of 0.93 and 0.89 for movie and video game respectively. We are in a process of conducting a user study with several sensory devices to explore combined physiological responses using IAPS data set as a stimulus.