Stress Detectection via Facial Emotion Recognition using YOLOv11model

In the modern era of fast-paced lifestyles and increasing psychological pressure, real-time stress identification has become a critical component of mental health monitoring systems. This research presents a novel approach to stress detection through facial expression analysis, leveraging Proximal Policy Optimization (PPO)–based Reinforcement Learning and self-attention mechanisms.We introduce a custom-labeled facial expression dataset with seven distinct emotional states-Bursted,Irritated,Anxious,Relaxed,Neutral,Broked,and Shocked to-accurately represent varying levels of psychological stress. A YOLOv11 model is employed for real-time facial region detection, ensuring efficient and high -speed localization in live video streams. The extracted facial features are then processed through a self-attention-enhanced PPO agent trained to classify stress leves based on subtle expression patterns.Experimental result demonstrate the robustness and accuracy of the proposed system in real-time environment, showing significant promise for integration into wearable devices,workplace monitoring systems,and tele-health applications.This work marks a significant step toward intelligent, non-invasive, and continuous mental state monitoring using cutting-edge deep reinforcement learning and real-time computer vision techniques.