An attention-based hybrid deep learning model for EEG emotion recognition

Emotion recognition based on electroencephalography (EEG) has received much attention in recent years, and there is more and more research on emotion recognition utilizing deep learning. It is difficult to extract more discriminative features for emotion recognition. To solve this problem, an attention-based hybrid deep learning model is proposed for EEG emotion recognition. The proposed approach extracts the critical feature information and achieves an excellent classification effect. To begin, the differential entropy features of EEG data are extracted and organized according to electrodeposition. Then, the convolutional encoder is used to encode the EEG signal and extract the spatial features, and the band attention mechanism is introduced to assign adaptive weights to different bands. Finally, a long short-term memory network is utilized to extract temporal features, and a time attention mechanism is used to obtain critical temporal information. The performance of the proposed model is analyzed on benchmark emotion databases such as DEAP and SEED for the classification task. The experimental results in terms of accuracy are 85.86% and 84.27% on DEAP datasets and 92.47% on SEED datasets. The experimental analysis shows that the proposed model can effectively recognize emotions and has a good classification performance.