Stress is a well known concept for all as everybody experiences stress sometimes, and many experience stress on a frequent basis. The experience of stress in itself is not necessarily a bad thing, however, when stress is experienced on a frequent basis and when one feels to have a lack of control on the experienced stress, it underlies many physical and psychological diseases.
To measure acute (present) stress, many different methods are used, such as the measurement of physiological measures (e.g. skin conductance and cardiac markers). Though reliable in accuracy, these measures require some type of interaction with a professional. Recently, with the increasingly popularity of smartwatches, these type of data have become more reliable. Yet, there is still a necessity to increase the accuracy to measure one’s stress levels remotely, also in individuals that are not wearing smart devices.
In recent literature, the possibility of stress detection through speech has been opted since the production of speech, even though it feels like a simple process to us, involves many different bodily processes and parts. Therefore, we are focusing on the development of accurate stress detection methods from spoken words to take the next step towards a method of remote stress assessments that has an unlimited range of applications, such as employee stress levels, strenuous/high-risk jobs, and relapse detection in psychological diseases.
For more information on this line of research, feel free to inquire with mitchel.kappen@ugent.be!
