Emotional Analysis and Recognition Based on EEG Brain Network

Emotions are closely linked to the life of human beings. Studying the difference of brain mechanism and brain function of human emotional expression helps to improve the human-computer interaction ability and realize the automatic diagnosis of related mental diseases. However, the emotion recognition rate is generally low. In this study, phase lock is used to construct EEG functional brain network based on static connection and dynamic connection under the condition of audiovisual stimuli of different core emotions. The network features are extracted for emotion recognition. The results demonstrate that the use of non-singular local attributes can effectively improve the rate of emotional recognition compared to the traditional use of global attributes of the network. At the same time, the dynamic brain network is introduced for the deficiency of the brain network analysis method based on a static connection. Constructing a dynamic connection-based EEG brain network whose network characteristics can further improve the emotional recognition rate. This conclusion provides some experimental basis and methods for relating emotion recognition based on brain network.