Hierarchical deep neural network for mental stress state detection using IoT based biomarkers

Affective state recognition at an early stage can help in mood stabilization, stress and depression management for mental well-being. Pro-active remote healthcare warrants the use of various biomarkers to detect affective individual by evaluating daily activities. With easy accessibility IoT-based sensors healthcare, observable quantifiable characteristics our body, physiological changes body be measured tracked using wearable devices. This work puts forward a model detection sensor-based bio-signals. A multi-level deep neural network with hierarchical learning capabilities convolution is proposed. Multivariate time-series data consisting both wrist-based chest-based sensor bio-signals trained hierarchy networks generate high-level features each bio-signal feature. model-level fusion strategy proposed combine into one unified representation classify states three categories as baseline, amusement. The evaluated on WESAD benchmark dataset health compares favourably state-of-the-art approaches giving superlative performance accuracy 87.7%.