English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

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

Learning to detect RFI in radio astronomy without seeing it