AI and the World's Hard Problems · Astronomy & space · arXiv · 2026-10-20
A targeted machine learning approach for detecting diffuse radio emission with Astronomaly: Protege
Diffuse radio emission in galaxy clusters is hard to detect because of its low surface brightness and contamination from compact sources and imaging artefacts. The authors extracted features with the self-supervised BYOL algorithm and used the Astronomaly: Protege anomaly detection framework with active learning to rank candidates in the MeerKAT Galaxy Cluster Legacy Survey. Of the top 100 sources, 98% showed diffuse characteristics and 53% were confirmed as cluster-related emission, with minimal human labelling.