Design of an Emotion Care System for the Elderly Based on Precisely Detecting Emotion States

With the global fertility rate in decline, the issue of an aging population grows ever more pressing, especially in developing nations where many elderly live alone, often deprived of familial emotional support and timely care. Addressing this, our study centers on emotional well-being and introduces a system capable of detecting and regulating the emotions of the elderly promptly. Utilizing sensors to capture facial imagery and vocal information, we employ a multimodal emotion recognition model for real-time analysis, enabling multi-channel emotional regulation. We validate the necessity for an elderly-specific dataset and confirm the reliability of our model's emotional recognition accuracy. Additionally, we demonstrate the effectiveness of our regulation strategies through empirical research. This study aims to enhance the quality of life for the elderly in their twilight years.