Subject-Aware Contrastive Learning for Biosignals

TL;DR

This paper introduces a novel self-supervised contrastive learning approach for biosignals that incorporates subject-aware mechanisms and data augmentation to improve performance with limited labeled data and subjects.

Contribution

It proposes a new contrastive learning framework with subject-specific loss and adversarial training for biosignals, addressing intersubject variability and data scarcity.

Findings

Subject-invariance enhances representation quality.

Subject-specific loss boosts fine-tuning performance.

Competitive results with supervised methods on biosignal tasks.

Abstract

Datasets for biosignals, such as electroencephalogram (EEG) and electrocardiogram (ECG), often have noisy labels and have limited number of subjects (<100). To handle these challenges, we propose a self-supervised approach based on contrastive learning to model biosignals with a reduced reliance on labeled data and with fewer subjects. In this regime of limited labels and subjects, intersubject variability negatively impacts model performance. Thus, we introduce subject-aware learning through (1) a subject-specific contrastive loss, and (2) an adversarial training to promote subject-invariance during the self-supervised learning. We also develop a number of time-series data augmentation techniques to be used with the contrastive loss for biosignals. Our method is evaluated on publicly available datasets of two different biosignals with different tasks: EEG decoding and ECG anomaly detection. The embeddings learned using self-supervision yield competitive classification results compared to entirely supervised methods. We show that subject-invariance improves representation quality for these tasks, and observe that subject-specific loss increases performance when fine-tuning with supervised labels.