GMSS: Graph-Based Multi-Task Self-Supervised Learning for EEG Emotion Recognition

TL;DR

This paper introduces GMSS, a graph-based multi-task self-supervised learning model that improves EEG emotion recognition by learning more general and discriminative features through multiple auxiliary tasks.

Contribution

The paper proposes a novel multi-task self-supervised learning framework for EEG emotion recognition that integrates spatial, frequency, and contrastive tasks to enhance feature generalization.

Findings

GMSS outperforms several existing methods on multiple datasets.

The model captures intrinsic spatial and frequency relationships in EEG signals.

Contrastive learning regularizes features for better generalization.

Abstract

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.