Radio Galaxy Zoo: Giant Radio Galaxy Classification using Multi-Domain Deep Learning

TL;DR

This paper develops a multi-domain deep learning approach to identify rare giant radio galaxies from large survey datasets, improving classification accuracy by integrating multi-resolution survey data and redshift information.

Contribution

The work introduces a multi-branch CNN model that learns from multiple survey inputs and redshift data, enhancing the detection of giant radio galaxies.

Findings

Multi-domain CNN reduces misclassification by 39% compared to single domain models.

Inclusion of redshift data moderately improves classification accuracy.

The approach is suitable for large-scale upcoming radio surveys like SKA.

Abstract

In this work, we explore the potential of multi-domain multi-branch convolutional neural networks (CNNs) for identifying comparatively rare giant radio galaxies from large volumes of survey data, such as those expected for new-generation radio telescopes like the SKA and its precursors. The approach presented here allows models to learn jointly from multiple survey inputs, in this case NVSS and FIRST, as well as incorporating numerical redshift information. We find that the inclusion of multi-resolution survey data results in correction of 39% of the misclassifications seen from equivalent single domain networks for the classification problem considered in this work. We also show that the inclusion of redshift information can moderately improve the classification of giant radio galaxies.