Design and Implementation of an Emotion-Aware Learning System Based on Facial Expression Recognition

This study introduces an emotion perception learning system that detects students' emotions through real-time recognition and analysis of facial expressions. This system is designed to assist teachers in implementing personalised education models and making dynamic teaching adjustments. The system is capable of recognising seven common emotional states: happy, sad, angry, surprised, disgusted, neutral and fearful. Its overall recognition accuracy is 82.5%. The system demonstrates particular efficacy in recognising and detecting emotions such as happiness and surprise. The experimental results demonstrate that positive emotions can enhance learning efficiency by 12%. Furthermore, timely intervention and feedback from the system can assist 85% of students in adjusting their attention following a negative emotional state. Furthermore, the system is suitable for deployment in large-scale classroom settings, offering rapid response times (average latency: 1.8 seconds), robust scalability, and reliable long-term operation. The potential issues pertaining to privacy and ethics are addressed through the implementation of data de-identification, encryption, and informed consent. Future developments will concentrate on enhancing the accuracy of recognition in low-light conditions and on refining the modes and types of feedback in order to adapt them to different educational scenarios. This research demonstrates the potential of affective sensing systems to facilitate the enhancement of personalised learning and classroom management.