Morphological Radio Galaxy Classification with a Fourier Convolutional Neural Network

Morphological classification of observed radio sources has become fundamental in studies related to the formation and evolution of galaxies and is therefore important to modern astronomy as a whole. Recently, Convolutional Neural Networks (CNNs) have enabled researchers to implement network architectures that are capable of classifying radio galaxies with high accuracies. However, the spatial correlations performed in these networks are computationally expensive. We therefore propose to replace the spatial correlations in a CNN with elementwise multiplications in the frequency domain to reduce the computational complexity, thus reducing the classification time.