Affective computing for stress, anxiety and cybersickness detection in virtual reality

The prevalence of stress and anxiety has increased dramatically in recent decades, especially with the global COVID-19 pandemic. In parallel, effective ways of objectively assessing and quantifying these conditions have continued to be explored. Affective computing is one such technique that has gained popularity recently, using physiological signals to interpret, and infer human emotion. Additionally, virtual reality (VR) is a rapidly developing tool with promising advancements in the health sciences. Given the emergence of new unobtrusive wearables and biosensors-instrumented VR headsets, the combined use of VR and affective computing has enabled the development of new immersive applications to objectively evaluate stress and anxiety. In this paper, we examine various affective computing methods that have been combined with VR with the goal of quantitatively measuring stress and anxiety levels. Additionally, we explore how affective computing has been used in the assessment of cybersickness. In particular, we surveyed current VR studies and summarized the most common physiological measurements used to characterize stress, anxiety, and cybersickness. Methods monitoring heart rate, skin conductance, muscle movement, and brain activity are described. We highlight the current challenges and propose opportunities for future research directions.Keywords: Stress, Anxiety, Affective Computing, Physiological Measurements, Virtual Reality (VR), Cybersickness