Learning to detect RFI in radio astronomy without seeing it

TL;DR

This paper introduces a novel RFI detection method in radio astronomy that learns from uncontaminated data using autoencoding and novelty detection, avoiding the need for extensive manual labeling.

Contribution

The authors propose a new approach using Nearest-Latent-Neighbours with autoencoders to detect RFI without requiring large labeled datasets, outperforming current methods on simulated data.

Findings

Outperforms state-of-the-art by ~1% AUROC and 3% AUPRC on simulated HERA data.

Achieves 4% improvement in AUROC and AUPRC on real LOFAR data without manual labels.

Provides a small expert-labelled dataset for evaluation of RFI detection methods.

Abstract

Radio Frequency Interference (RFI) corrupts astronomical measurements, thus affecting the performance of radio telescopes. To address this problem, supervised segmentation models have been proposed as candidate solutions to RFI detection. However, the unavailability of large labelled datasets, due to the prohibitive cost of annotating, makes these solutions unusable. To solve these shortcomings, we focus on the inverse problem; training models on only uncontaminated emissions thereby learning to discriminate RFI from all known astronomical signals and system noise. We use Nearest-Latent-Neighbours (NLN) - an algorithm that utilises both the reconstructions and latent distances to the nearest-neighbours in the latent space of generative autoencoding models for novelty detection. The uncontaminated regions are selected using weak-labels in the form of RFI flags (generated by classical RFI flagging methods) available from most radio astronomical data archives at no additional cost. We evaluate performance on two independent datasets, one simulated from the HERA telescope and another consisting of real observations from LOFAR telescope. Additionally, we provide a small expert-labelled LOFAR dataset (i.e., strong labels) for evaluation of our and other methods. Performance is measured using AUROC, AUPRC and the maximum F1-score for a fixed threshold. For the simulated data we outperform the current state-of-the-art by approximately 1% in AUROC and 3% in AUPRC for the HERA dataset. Furthermore, our algorithm offers both a 4% increase in AUROC and AUPRC at a cost of a degradation in F1-score performance for the LOFAR dataset, without any manual labelling.