Identifying of Different Emotional Stimulation Modes from High-Density Frontal EEG for Emotion Recognition

Research on emotion recognition using electroencephalography (EEG) mainly concentrates on improving model classification accuracy based the whole brain channels of EEG signals with the preset emotional stimulation scene. However, the emotional triggers in practical application scenarios are more random and uncertain. This study innovatively used High-Density frontal EEG signals (HD-fEEG) to classify three basic emotions and explore the performance of emotion recognition system under different stimulus duration. Experimental results show that the average classification accuracy with continuous single emotional stimulation paradigm among all participants obviously surpassed the random presentation of emotional trials. Our research conclusion will provide insightful guidance for improving the performance of emotion recognition systems in real-world applications.