EEG Emotion Recognition Using Dynamical Graph Convolutional Neural Networks

By Tengfei Song, Wenming Zheng, Peng Song, Zhen Cui. IEEE Transactions on Affective Computing, 2020

1,382 citations recorded by the source corpus.

In this paper, a multichannel EEG emotion recognition method based on a novel dynamical graph convolutional neural networks (DGCNN) is proposed. The basic idea of the proposed EEG emotion recognition method is to use a graph to model the multichannel EEG features and then perform EEG emotion classification based on this model. Different from the traditional graph convolutional neural networks (GCNN) methods, the proposed DGCNN method can dynamically learn the intrinsic relationship between different electroencephalogram (EEG) channels, represented by an adjacency matrix, via training a neural network so as to benefit for more discriminative EEG feature extraction. Then, the learned adjacency matrix is used to learn more discriminative features for improving the EEG emotion recognition. We conduct extensive experiments on the SJTU emotion EEG dataset (SEED) and DREAMER dataset. The experimental results demonstrate that the proposed method achieves better recognition performance than the state-of-the-art methods, in which the average recognition accuracy of 90.4 percent is achieved for subject dependent experiment while 79.95 percent for subject independent cross-validation one on the SEED database, and the average accuracies of 86.23, 84.54 and 85.02 percent are respectively obtained for valence, arousal and dominance classifications on the DREAMER database.

Subfields: Cognitive Neuroscience, Experimental and Cognitive Psychology, Human-Computer Interaction

View Paper: https://www.semanticscholar.org/paper/33780d6c82a0c060a9a35fd07effbd2fbb3e5b82

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

1. Dongrui Wu
2. Wei-Long Zheng
3. Chunyan Miao
4. Di Wang
5. Peixiang Zhong
6. Wei Liu
7. Zhen Cui
8. Wenming Zheng

In Experimental and Cognitive Psychology

1. Wei-Long Zheng
2. Chunyan Miao
3. Di Wang
4. Peixiang Zhong
5. Dongrui Wu
6. Wei Liu
7. Zhen Cui
8. Wenming Zheng

In Human-Computer Interaction

1. Dongrui Wu
2. Chin-Teng Lin
3. Alireza Jolfaei
4. Peng Xu
5. Xiaotong Gu
6. Zehong Cao
7. Tzyy-Ping Jung
8. Wei-Long Zheng

Citations in the corpus

• Transfer Learning for EEG-Based Brain–Computer Interfaces: A Review of Progress Made Since 2016 (2020, positive): “[101] proposed a dynamical graph convolutional neural network (DGCNN) for subject-dependent and subjectindependent emotion classification.”
• A Bi-Hemisphere Domain Adversarial Neural Network Model for EEG Emotion Recognition (2021, methodology): “…(CCA) [39], group sparse canonical correlation analysis [11], deep believe network [13], graph regularization sparse linear regression [12], graph convolutional neural network (GCNN) [40] and dynamical graph convolutional neural network (DGCNN) [14] to conduct the same experiment, respectively.”
• EEG Conformer: Convolutional Transformer for EEG Decoding and Visualization (2022, methodology): “Besides, graph-based neural networks learning the intrinsic relationship among different EEG channels such as dynamical graph convolutional neural network (DGCNN) [41] and regularized graph neural network (RGNN) [43] are also included for comparison.”
• Comparing Recognition Performance and Robustness of Multimodal Deep Learning Models for Multimodal Emotion Recognition (2021, methodology): “[53] proposed DGCNN and obtained a classification accuracy of 90.”
• A Review of Emotion Recognition Using Physiological Signals (2018, background): “The authors of [137] proposes novel dynamical graph convolutional neural networks (DGCNN) and obtained the highest recognition accuracy of 90.”
• EEG-Based Brain-Computer Interfaces (BCIs): A Survey of Recent Studies on Signal Sensing Technologies and Computational Intelligence Approaches and Their Applications (2020, methodology): “[119], the presented DGCNNmethod uses a graph to model EEG features by learning intrinsic correlations between each EEG channel to produce an adjacency matrix, which is then applied to learn more discriminative features.”
• EEG-Based Emotion Recognition Using Regularized Graph Neural Networks (2019, methodology): “Graph neural networks (GNN) has been applied in [12] to capture inter-channel relations using an adjacency matrix.”
• EEG Based Emotion Recognition: A Tutorial and Review (2022, methodology): “[137] proposed a dynamical graph convolutional neural networks (DGCNN)”
• An Efficient LSTM Network for Emotion Recognition From Multichannel EEG Signals (2020, methodology): “Moreover, the DGCNN method is specially designed as an improved version of GCNN for multichannel EEG signal analysis [14].”
• Large Brain Model for Learning Generic Representations with Tremendous EEG Data in BCI (2024, methodology): “• SEED Series (Zheng & Lu, 2015; Zheng et al., 2018; Liu et al., 2022): (von Weltin et al., 2017): This is another subset of TUEG that contains annotations of slowing events (23 channels, 256 Hz).”
• EEG-Based Emotion Recognition via Channel-Wise Attention and Self Attention (2023, methodology): “Compared to the three deep learning methods (Conti-CNN, CRAM, and GCNN), the experimental results indicate that the proposed ACRNN improved the average recognition accuracy by 10, 8 and 5 percent, respectively.”
• EEG-Based Emotion Recognition: A State-of-the-Art Review of Current Trends and Opportunities (2020, methodology): “[110] Negative, positive, and neutral (SEED).”
• A Novel Bi-Hemispheric Discrepancy Model for EEG Emotion Recognition (2019, methodology): “[26] used the dynamic graph convolution neural network (DGCNN) to model the multichannel EEG features and they also presented a novel attention-long”