Subject Representation Learning from EEG using Graph Convolutional Variational Autoencoders

TL;DR

This paper introduces GC-VASE, a graph convolutional variational autoencoder with contrastive learning and attention-based adapters, achieving state-of-the-art subject identification accuracy from EEG data with efficient adaptation to new subjects.

Contribution

The paper presents a novel EEG subject representation learning method combining graph convolutional VAEs, contrastive learning, and attention-based adapters for efficient adaptation.

Findings

Achieves 89.81% accuracy on ERP-Core dataset

Improves to 90.31% accuracy with fine-tuning

Outperforms existing deep learning methods

Abstract

We propose GC-VASE, a graph convolutional-based variational autoencoder that leverages contrastive learning for subject representation learning from EEG data. Our method successfully learns robust subject-specific latent representations using the split-latent space architecture tailored for subject identification. To enhance the model's adaptability to unseen subjects without extensive retraining, we introduce an attention-based adapter network for fine-tuning, which reduces the computational cost of adapting the model to new subjects. Our method significantly outperforms other deep learning approaches, achieving state-of-the-art results with a subject balanced accuracy of 89.81% on the ERP-Core dataset and 70.85% on the SleepEDFx-20 dataset. After subject adaptive fine-tuning using adapters and attention layers, GC-VASE further improves the subject balanced accuracy to 90.31% on ERP-Core. Additionally, we perform a detailed ablation study to highlight the impact of the key components of our method.