Exploring Real-Life Psychophysiological States with Self-Supervised Learning

Affective experiences play an essential role in human life, yet their physiological underpinnings are not well studied in everyday settings. Therefore, moving beyond traditional laboratory studies is necessary. This research aims to explore psychophysiological states associated with affect in real-life contexts by applying self-supervised learning (SSL) and unsupervised clustering techniques to physiological signals collected through wearables. Preliminary results demonstrate the potential of SSL models in capturing meaningful patterns in physiological data.