AI-Based Emotion State Monitoring: Methods, Applications, and Limitations

Abstract

Emotional state monitoring has become an important research topic of artificial intelligence, which has applications in mental health, human-computer interaction and intelligent systems. This paper provides a problem-oriented review of emotion monitoring technology based on AI. Firstly, the basic challenges of emotion recognition are analyzed, including subjectivity, context dependence and the complexity of multi-mode. Next, the main application scenarios, such as mental health monitoring and intelligent interaction, are reviewed. Finally, the limitations of current methods-including data quality, interpretability, privacy and moral hazard-are critically analyzed, and the future research direction is put forward. This review shows that the current emotion recognition system is still mainly based on probabilistic reasoning, rather than real cognitive understanding, emphasizing the need to balance technological progress and moral responsibility.