Real-Life Stress Level Monitoring using Smart Bands in the Light of Contextual Information

An automatic stress detection system that uses unobtrusive\r\nsmart bands will contribute to human health and wellbeing\r\nby alleviating the effects of high stress levels. However, there\r\nare a number of challenges for detecting stress in unrestricted\r\ndaily life which results in lower performances of such systems\r\nwhen compared to semi-restricted and laboratory environment\r\nstudies. The addition of contextual information such as physical\r\nactivity level, activity type and weather to the physiological\r\nsignals can improve the classification accuracies of these systems.\r\nWe developed an automatic stress detection system that employs\r\nsmart bands for physiological data collection. In this study, we\r\nmonitored the stress levels of 16 participants of an EU project\r\ntraining every day throughout the eight days long event by\r\nusing our system. We collected 1440 hours of physiological data\r\nand 2780 self-report questions from the participants who are\r\nfrom diverse countries. The project midterm presentations (see\r\nFigure 3) in front of a jury at the end of the event were the\r\nsource of significant real stress. Different types of contextual\r\ninformation, along with the physiological data, were recorded to\r\ndetermine the perceived stress levels of individuals. We further\r\nanalyze the physiological signals in this event to infer long term\r\nperceived stress levels which we obtained from baseline PSS-\r\n14 questionnaires. Session-based, daily and long-term perceived\r\nstress levels could be identified by using the proposed system\r\nsuccessfully.