Stress generation and non-intrusive measurement in virtual environments using eye tracking

Abstract

In real life, it is well understood how stress can be induced and how it is measured. While virtual reality (VR) applications can resemble such stress inducers, it is still an open question if and how stress can be measured in a non-intrusive way during VR exposure. Usually, the quality of VR applications is estimated by user acceptance in the form of presence. Presence itself describes the individual's acceptance of a virtual environment as real and is measured by specific questionnaires. Accordingly, it is expected that stress strongly affects this presence and thus also the quality assessment. Consequently, identifying the stress level of a VR user may enable content creators to engage users more immersively by adjusting the virtual environment to the measured stress. In this paper, we thus propose to use a commercially available eye tracking device to detect stress while users are exploring a virtual environment. We describe a user study in which a VR task was implemented to induce stress, while users' pupil diameter and pulse were measured and evaluated against a self-reported stress level. The results show a statistically significant correlation between self-reported stress and users' pupil dilation and pulse, indicating that stress measurements can indeed be conducted during the use of a head-mounted display. If this indication can be successfully proven in a larger scope, it will open up a new era of affective VR applications using individual and dynamic adjustments in the virtual environment.

Human Stress Recognition from Facial Thermal-Based Signature: A Literature Survey

“…Modal Feature [33,34] Eyes activity Pupil Dilation Gaze Movement Blinking rate [35] Key stroke Key stroke pattern and dilation [36] Interaction on social media Detect in user post on social media [37][38][39] Speech signal Pitch Jitter Energy Speaking rate Length of pauses [40][41][42] Facial expression Local binary patterns (LBP-TOP) 3D histogram of oriented gradients (3DHOG) Weighted random forest (WRF) 3D scale-invariant feature transform (3DSIFT) (Continued) Many studies have explored the possibility of stress detection with facial skin temperature [45][46][47]. Researchers also investigated other modalities such as pupil dilation, breathing pattern, behavior pattern, keystroke pattern, and social media activity.…”

Human factors evaluation in VR-based shunting training

“…3), and they can be gathered on a by-task basis to visualize trends and enable day-to-day performance comparison. The fatigue and stress recognition algorithms are implemented based on the stateof-art methods published in [17][18][19][20], respectively.…”

A Review of Emotion Recognition Methods From Keystroke, Mouse, and Touchscreen Dynamics

“…They can be generally classified into three categories: a) Use external body signals for emotion recognition, including facial expression, body gestures, gait, eyetracking, etc. These signals can be easily noticed by others but not always reflect one's real emotional states [15][16][17][18][19][20][21][22][23]. b) Use internal physiological signals such as heart rate, sphygmic, skin conductance, blood pressure, Electroencephalography (EEG), etc.…”