Quantifying place-based emotional arousal: A multi-modal wearable sensing approach to understanding human-environment interaction

Understanding the intricate interplay between human emotions and the unique characteristics of place is crucial for designing emotionally resonant spaces and enhancing the experiential value of environments. This study introduces a novel approach to quantifying place-based emotional arousal by leveraging multi-modal wearable sensing. We conducted field experiments across diverse urban settings, collecting participants’ physiological responses, ambient audiovisual information, and GPS trajectories to capture the dynamic interplay between individuals and their surroundings. Our dataset links emotional arousal with perceptual audiovisual entropy features extracted from the environment. Employing random forests and SHAP analysis, we developed interpretable predictive models that reveal how specific sensory attributes of place contribute to emotional arousal. Our findings demonstrate that low-dimensional perceptual audiovisual entropy explains a substantial portion of emotional arousal, with distinct patterns emerging across different place types. This research offers a quantitative framework for investigating human-environment interaction, advancing our understanding of the affective dimension of place and informing the design of environments that foster emotional well-being. • A novel framework quantifies place-based emotional arousal and audiovisual entropy using wearable sensing. • The result indicates that higher audiovisual entropy is linked to increased emotional arousal. • RF model (R 2 > 0.50) effectively predicts emotional arousal based on audiovisual entropy across different places.