Emotion Classification from Multi-Band Electroencephalogram Data Using Dynamic Simplifying Graph Convolutional Network and Channel Style Recalibration Module

Abstract

Because of its ability to objectively reflect people’s emotional states, electroencephalogram (EEG) has been attracting increasing research attention for emotion classification. The classification method based on spatial-domain analysis is one of the research hotspots. However, most previous studies ignored the complementarity of information between different frequency bands, and the information in a single frequency band is not fully mined, which increases the computational time and the difficulty of improving classification accuracy. To address the above problems, this study proposes an emotion classification method based on dynamic simplifying graph convolutional (SGC) networks and a style recalibration module (SRM) for channels, termed SGC-SRM, with multi-band EEG data as input. Specifically, first, the graph structure is constructed using the differential entropy characteristics of each sub-band and the internal relationship between different channels is dynamically learned through SGC networks. Second, a convolution layer based on the SRM is introduced to recalibrate channel features to extract more emotion-related features. Third, the extracted sub-band features are fused at the feature level and classified. In addition, to reduce the redundant information between EEG channels and the computational time, (1) we adopt only 12 channels that are suitable for emotion classification to optimize the recognition algorithm, which can save approximately 90.5% of the time cost compared with using all channels; (2) we adopt information in the θ, α, β, and γ bands, consequently saving 23.3% of the time consumed compared with that in the full bands while maintaining almost the same level of classification accuracy. Finally, a subject-independent experiment is conducted on the public SEED dataset using the leave-one-subject-out cross-validation strategy. According to experimental results, SGC-SRM improves classification accuracy by 5.51–15.43% compared with existing methods.

EEG Emotion Recognition Network Based on Attention and Spatiotemporal Convolution

“…The recognition rates for the negative, neutral, and positive emotions were 93.78%, 95.56%, and 96.44%, respectively. Consistent with the results of previous studies, the positive and neutral emotions were easier to recognize than the negative emotions [ 28 ]. Note that negative and neutral emotions are more easily confused.…”

“…Here, we set the input features as a single frequency band and a combination frequency band. Here, many studies use (θ, α, β, and γ), and we conducted experiments on these four frequency bands [ 28 ] and provide corresponding experimental results. As shown in Table 2 , the proposed model performs well on the δ, θ, α, β, and γ frequency bands.…”

Optimizing 1D-CNN-Based Emotion Recognition Process through Channel and Feature Selection from EEG Signals

“…However, owing to the high non-stationarity of EEG signals, extracting such features is challenging and requires expert knowledge. Recent studies have explored multivariate statistical analysis techniques in the frequency, time-frequency, and nonlinear domains that effectively represent EEG characteristics [ 7 , 8 , 9 ]. For instance, Zheng et al [ 7 ] evaluated different feature extraction methods and achieved the best average accuracy using a discriminative graph-regularized Extreme Learning Machine with differential entropy features.…”

“…For instance, Zheng et al [ 7 ] evaluated different feature extraction methods and achieved the best average accuracy using a discriminative graph-regularized Extreme Learning Machine with differential entropy features. Zhu et al [ 8 ] constructed a graph structure based on differential entropy characteristics, learned channel relationships through dynamic simplifying graph convolutional networks, and recalibrated channel features using the style recalibration module. Through the fusion and classification of sub-band features, the method achieves improved classification accuracy compared to existing methods.…”

EEG-Based Emotion Recognition with Consideration of Individual Difference

“…Li et al [ 17 ] introduced a multidimensional approach based on the continuous wavelet transform and the Clough–Tocher interpolation algorithm for processing motor intention electroencephalography (MI-EEG) signals combined with a multilevel and multiscale feature fusion convolutional neural network (MLMSFFCNN) for recognition. Zhu et al [ 18 ] proposed an emotion recognition method considering multi-band EEG data inputs based on a dynamic Simplified Graph Convolution (SGC) network and a channel-style recalibration module. Zhang et al [ 19 ] proposed the idea of assigning channel weight ratios to the channels that are more strongly correlated with emotion.…”