A Hierarchical Bidirectional GRU Model With Attention for EEG-Based Emotion Classification

Abstract

In this paper, we propose a hierarchical bidirectional Gated Recurrent Unit (GRU) network with attention for human emotion classification from continues electroencephalogram (EEG) signals. The structure of the model mirrors the hierarchical structure of EEG signals, and the attention mechanism is used at two levels of EEG samples and epochs. By paying different levels of attention to content with different importance, the model can learn more significant feature representation of EEG sequence which highlights the contribution of important samples and epochs to its emotional categories. We conduct the cross-subject emotion classification experiments on DEAP data set to evaluate the model performance. The experimental results show that in valence and arousal dimensions, our model on 1-s segmented EEG sequences outperforms the best deep baseline LSTM model by 4.2% and 4.6%, and outperforms the best shallow baseline model by 11.7% and 12% respectively. Moreover, with increase of the epoch's length of EEG sequences, our model shows more robust classification performance than baseline models, which demonstrates that the proposed model can effectively reduce the impact of long-term non-stationarity of EEG sequences and improve the accuracy and robustness of EEG-based emotion classification.

EEG-Based Emotion Recognition Using Logistic Regression with Gaussian Kernel and Laplacian Prior and Investigation of Critical Frequency Bands

“…The classification accuracies for valence and arousal were over 85% using LSTM-RNN by Alhagry et al [77], and over 87% using 3D-CNN by Salama et al [78]. More recently, Chen et al [33,34] have researched a lot on the combination of DL models and various features. As tabulated in Table 5, computer vision CNN (CVCNN), global spatial filter CNN (GSCNN), and global space local time filter (GSLTCNN) [33] presented obvious improvements with concatenating PSD, raw EEG features, and normalized EEG signals.…”

“…As tabulated in Table 5, computer vision CNN (CVCNN), global spatial filter CNN (GSCNN), and global space local time filter (GSLTCNN) [33] presented obvious improvements with concatenating PSD, raw EEG features, and normalized EEG signals. In [34], the proposed hierarchical bidirectional gated recurrent unit (H-ATT-BGRU) network performed better on raw EEG signals than CNN and LSTM, and the obtained accuracies in valence and arousal dimensions were 67.9% and 66.5% for 2-class cross-subject emotion recognition. For more details about the DL architectures applied in the DEAP data, readers may refer to the literature [33,34,[75][76][77][78].…”

“…The obtained emotion classification accuracies are higher than the traditionally best bagging tree (BT) classifier. The study [34] proposed a hierarchical bidirectional gated recurrent unit (GRU) network with an attention mechanism. The proposed scheme learned more significant representation from EEG sequences and the accuracies obtained on cross-subject emotion classification task outperformed the long short time memory (LSTM) network by 4.2% and 4.6% in valence and arousal dimensions, respectively.…”

Hybrid Network with Attention Mechanism for Detection and Location of Myocardial Infarction Based on 12-Lead Electrocardiogram Signals

“…In processing biomedical signals, BiGRU has been successfully applied for human emotion classification through continuous electroencephalogram signals [41], and human identification through ECG based biometrics [42]. ECG signal is a typical kind of time series data, and LSTM has been effectively applied in MI diagnosis [21][22][23].…”

Improved Deep Feature Learning by Synchronization Measurements for Multi-Channel EEG Emotion Recognition

“…Table 4 presents the details of the comparison methods and emotion recognition results. Among them, features extracted from the central nervous system (CNS) were used in reference [42] and hierarchical bidirectional Gated Recurrent Unit network (H-ATT-BGRU) were used in reference [51].…”