EMO-Care: Emotional Interaction System Based on Multimodal Fusion and Edge Intelligence for Emotional Care

In recent years, driven by the new generation of artificial intelligence technology, emotional computing has garnered significant attention and achieved substantial progress. Multimodal emotional computing based on technologies such as deep learning, computer vision, wearable computing, multimodal machine learning, and reinforcement learning has emerged as a research focus. Emotional computing research is shifting from the cognition of human emotions to a novel stage of emotional interaction. It becomes feasible to construct multimodal emotional interaction application systems with the ability of emotional care. This paper proposes a multimodal emotional care system based on wearable computing and reinforcement learning, introduces the overall architecture and core components of the system, and discusses the problems and solutions faced by the system, which has important value for how to build a dynamic, real-time and highly personalized emotional interaction application based on the user's current emotional state.