Women Stress Detection System using Neural Networks

Stress detection is crucial due to its impact on mental and physical health. This paper proposes a non-invasive method for real-time stress detection using multimodal physiological data, including accelerometer readings, temperature, Galvanic Skin Response (GSR), and heart rate. By integrating these diverse signals, we develop a neural network-based model capable of accurately classifying stress levels into low, medium, and high categories. Our approach involves data collection, preprocessing, and feature extraction, followed by training a feedforward neural network (FNN) to detect stress patterns. The dataset includes physiological signals recorded over a specific period, ensuring a robust and diverse representation of stress responses. Compared to traditional models like SVM, Random Forest, and KNN(K-Nearest Neighbors), our feedforward neural network(FNN) achieve better accuracy, demonstrating its superior performance in stress classification. Additionally, we have developed a user-friendly web interface using Flask, HTML, CSS, and JavaScript for real-time stress monitoring. This system enables individuals to track their stress levels efficiently, promoting early intervention and better stress management. Our research contributes to the advancement of AI-driven stress detection, offering a scalable and effective solution for mental health monitoring.