Navigating Mental Health: Emotion and Personality-Based Prediction System

Abstract

Easily available, customized and efficient mental health support services require immediate focus as seen by the rise in issues related to mental health. People looking for quick measures face challenges due to the difficult, subjective or lacking nature of standard procedures for monitoring for mental health issues. Awareness of an individual's mental condition requires an awareness of their mental health and personality features but these elements are rarely integrated into a coherent framework for continuous monitoring and management. This study fills this gap by using cutting-edge machine learning algorithms to forecast mental health disorders based on personality trait analysis and real-time emotion recognition. With actual time capture of images using OpenCV the system uses a CNN to reliably recognize emotions including "Anger," "Contempt," "Disgust," "Fear," "Happy," "Neutral," "Sad" and "Surprised." It achieves an impressive 94% accuracy in training and 100% testing accuracy. Furthermore, personality qualities such as Reliable, Extraverted, Serious, Lively and Responsible are predicted using Logistic Regression. By combining these predictions, the system evaluates the person's mental health and offers useful information such as specific mental health advice and suggestions for relaxing or inspirational videos selected through the YouTube API. Real-time heath detection and personalized treatments are made possible by an interface that guarantees accessibility and promptness.