Stress Detection from Wearable Sensor Data Using Gramian Angular Fields and CNN

Stress is a body reaction that is one of the principal causes of many physical and mental disorders, including cardiovascular disease and depression. Developing robust methods for rapid and accurate stress detection plays an important role in improving people’s life quality and wellness. Prior research shows that analyzing physiological signals collected from wearable sensors is a reliable predictor of stress. For stress detection, methods based on machine learning techniques have been defined in the literature. However, they require hand-crafted features to be effective. Deep learning-based approaches overcome these limitations. In this work, we introduce STREDWES, a method for stress detection that analyzes biosignals obtained from wearable sensor data. STREDWES extracts signal fragments using a sliding windows approach and converts them into Gramian Angular Fields images. These images are then classified using a Convolutional Neural Network, a deep learning algorithm. We apply our method to a publicly available dataset. The analysis of the performance values shows that our method outperforms other state-of-the-art competitors.