AI Based Stress and Mental Health Monitoring System Using Chatbot, Speech and Facial Analysis

Abstract

The increase in academic pressure and workplace stress has led to higher levels of mental illness, including stress, anxiety, and depression. When stress is not handled and recognized early, there can be significant health issues, both mentally and physically. Traditional ways of evaluating mental health rely on self-reporting and clinical interviews, which can be very subjective and time-consuming and not always able to be monitored continuously or in real-time. Because of this, artificial intelligence (AI) has emerged as a viable option for providing automated assessments of mental health through technology. This paper discusses the current state of AI-based monitoring of mental health and stress and provides an overview of studies that have examined AI-based chatbots, audio signal analysis, and facial recognition. This systematic review of the literature also provides a summary of various machine learning and deep learning techniques used to detect patterns of stress using multimodal data. The conclusions drawn from this study suggest that using multiple data sources together improves the accuracy and robustness of AI-based systems compared to single-modality systems, making them more suitable for practical use in mental health settings.