StressNet: A Machine Learning-Based IoT Framework for Stress Detection

Academic research focuses intensely on stress detection methods because of growing mental health issues. Current stress monitoring systems use wearable sensors however these sensors might not be always available for use. The proposed research introduces StressNet as an IoT-based framework for real-time stress detection operated through machine learning methods. The WESAD (Wearable Stress and Affect Detection) dataset was used instead of genuine sensors because it delivers physiological and motion data collected through devices worn on the chest and wrist. The deep learning model accepts normalized heart rate and skin conductance along with respiration rate and temperature attributes that underwent a transformation stage for segmentation and noise filtering. StressNet surpasses previous benchmarks in the assessment against existing models because it increases IoT-based stress detection adaptability without reducing performance accuracy. Key-words: Stress detection, IoT-based framework, machine learning, StressNet, WESAD dataset, physiological signals, deep learning, real-time monitoring, heart rate, skin conductance, respiration rate, temperature normalization, noise filtering, stress classification