期刊
IEEE TRANSACTIONS ON AFFECTIVE COMPUTING
卷 14, 期 3, 页码 2512-2525出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TAFFC.2022.3170428
关键词
EEG emotion recognition; multi-task learning; self-supervised learning; graph neural network
Previous electroencephalogram (EEG) emotion recognition heavily relies on single-task learning, which may lead to overfitting and a lack of generalization. In this paper, the authors propose a graph-based multi-task self-supervised learning model (GMSS) for EEG emotion recognition. The proposed model integrates multiple self-supervised tasks, including spatial and frequency jigsaw puzzles, and contrastive learning, to learn more general representations and improve the accuracy of emotion recognition.
Previous electroencephalogram (EEG) emotion recognition relies on single-task learning, which may lead to overfitting and learned emotion features lacking generalization. In this paper, a graph-based multi-task self-supervised learning model (GMSS) for EEG emotion recognition is proposed. GMSS has the ability to learn more general representations by integrating multiple self-supervised tasks, including spatial and frequency jigsaw puzzle tasks, and contrastive learning tasks. By learning from multiple tasks simultaneously, GMSS can find a representation that captures all of the tasks thereby decreasing the chance of overfitting on the original task, i.e., emotion recognition task. In particular, the spatial jigsaw puzzle task aims to capture the intrinsic spatial relationships of different brain regions. Considering the importance of frequency information in EEG emotional signals, the goal of the frequency jigsaw puzzle task is to explore the crucial frequency bands for EEG emotion recognition. To further regularize the learned features and encourage the network to learn inherent representations, contrastive learning task is adopted in this work by mapping the transformed data into a common feature space. The performance of the proposed GMSS is compared with several popular unsupervised and supervised methods. Experiments on SEED, SEED-IV, and MPED datasets show that the proposed model has remarkable advantages in learning more discriminative and general features for EEG emotional signals.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据