flâneur — a map of the web's best reading

[2007.04871] Subject-Aware Contrastive Learning for Biosignals

arxiv.org · 663 words · saved by 1 readers

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.

[2007.04871] Subject-Aware Contrastive Learning for Biosignals Skip to main content arXiv is now an independent nonprofit! Learn more &times; Search arXiv Press Enter to search &middot; Advanced search --> Computer Science > Machine Learning arXiv:2007.04871 (cs) [Submitted on 30 Jun 2020] Title: Subject-Aware Contrastive Learning for Biosignals Authors: Joseph Y. Cheng , Hanlin Goh , Kaan Dogrusoz , Oncel Tuzel , Erdrin Azemi View a PDF of the paper titled Subject-Aware Contrastive Learning for Biosignals, by Joseph Y. Cheng and 4 other authors View PDF Abstract: Datasets for biosignals, such

Explore this link on the map →

related reading