Personalized Biofeedback for Affective Dynamics: A Virtual Reality Framework Based on Markov Modeling

Affective dynamics, the unfolding of emotional states over time, are central to understanding emotion but remain difficult to capture experimentally. This paper presents a novel framework integrating immersive virtual reality (VR), heart rate variability (HRV) biofeedback, and discrete-time Markov modeling to capture and enhance emotional flexibility, offering a trajectory-focused approach to both measuring and training affective dynamics. Participants experience twelve validated VR transitions based on Russell's circumplex, while HRV is recorded and analyzed offline to create individual affective dynamic profiles. Using these, a personalized VR biofeedback protocol is delivered via the Excite-O-Meter platform. This trajectory-focused approach offers a promising tool for advancing affective science and developing adaptive emotional training systems.