Next-Gen Stress Monitoring: Social Robot and AI Integration

In recent years, there has been a surge in the popularity of social robots in various fields, including healthcare, which has affected people's health over time. In today's fast-paced world, many people experience stress and other emotional conditions that can affect their well-being. Social robots are increasingly being developed to provide emotional support and companionship to individuals in need. Early identification of stress is essential in order to address and deal with it. Artificial intelligence can play an important role in the detection of stress using facial expressions. The objective of this study is to detect emotions such as stress and no stress using the facial expressions of the persons. To achieve this goal, we examine the hybrid performance of three AI libraries- VGGFace, DLib, and DeepFace. Furthermore, a humanoid robot, Nao, is integrated with AI libraries, which helps in capturing people's facial images. The stress and non-stress situations are simulated by playing a game and watching an entertaining video, respectively and labelled accordingly. These images are then processed by algorithms individually, with detected emotions labeled as stress and non-stress. The findings of this study report 93.589% accuracy in classifying emotions. In the future, the robot's suggestions will also be incorporated to promote relaxation and stress reduction.