Extract non-Gaussian Features in Gravitational Wave Observation Data Using Self-Supervised Learning

TL;DR

This paper introduces a self-supervised neural network approach to denoise gravitational wave signals in time series data, improving detection and analysis without waveform reliance.

Contribution

It presents a novel self-supervised denoising method using blind-spot neural networks for gravitational wave data, capable of extracting signals and features without prior waveform models.

Findings

Successfully denoised 38% of GW signals in H1 data

Denoised 49% of GW signals in L1 data

Potential to extract glitch features and pre-merger signals

Abstract

We propose a self-supervised learning model to denoise gravitational wave (GW) signals in the time series strain data without relying on waveform information. Denoising GW data is a crucial intermediate process for machine-learning-based data analysis techniques, as it can simplify the model for downstream tasks such as detections and parameter estimations. We use the blind-spot neural network and train it with whitened strain data with GW signals injected as both input data and target. Under the assumption of a Gaussian noise model, our model successfully denoises 38% of GW signals from binary black hole mergers in H1 data and 49% of signals in L1 data detected in the O1, O2, and O3 observation runs with an overlap greater than 0.5. We also test the model's potential to extract glitch features, loud inspiral compact binary coalescence signals a few seconds before the merger, and unseen CCSN signals during training.