Personalized Stress Detection from Wearable Physiological Signals via Digital Twin-Inspired Subject Embeddings and Recurrent Neural Networks

Abstract

This paper presents a DT-GRU framework aimed at personalized stress detection in wearable sensing systems. With the rising prevalence of psychological stress and its adverse health effects, there is an urgent need for reliable and automated detection methods that move beyond traditional self-report tools. Current systems often fail to account for individual physiological variability in stress responses, leading to suboptimal performance across diverse populations. To address this challenge, we introduce a trainable subject-embedding module that captures each individual's unique physiological traits, thereby forming a latent DT. This personalized embedding is integrated with GRU features in the model's inference pathway, enabling the framework to utilize both temporal dynamics and subject identity. The DT-GRU framework is rigorously evaluated on the WESAD dataset, incorporating various validation techniques, including subject-independent train-test splits and Leave-One-Subject-Out (LOSO) cross-validation. Results demonstrate significant improvements in classification accuracy and generalizability, highlighting the potential of this approach for deployment in real-world stress monitoring applications.