Radio Galaxy Morphology Generation Using DNN Autoencoder and Gaussian Mixture Models

TL;DR

This paper presents a novel framework combining deep neural network autoencoders and Gaussian mixture models to generate realistic radio galaxy morphologies, aiding the study of active galactic nuclei and black hole evolution.

Contribution

It introduces a new method for generating radio galaxy images using autoencoders and GMMs, improving the realism and diversity of simulated galaxy morphologies.

Findings

High efficiency and performance demonstrated on real radio galaxy images

Effective generation of FRI and FRII galaxy morphologies

Analysis of feature vector length and training techniques

Abstract

The morphology of a radio galaxy is highly affected by its central active galactic nuclei (AGN), which is studied to reveal the evolution of the super massive black hole (SMBH). In this work, we propose a morphology generation framework for two typical radio galaxies namely Fanaroff-Riley type-I (FRI) and type-II (FRII) with deep neural network based autoencoder (DNNAE) and Gaussian mixture models (GMMs). The encoder and decoder subnets in the DNNAE are symmetric aside a fully-connected layer namely code layer hosting the extracted feature vectors. By randomly generating the feature vectors later with a three-component Gaussian Mixture models, new FRI or FRII radio galaxy morphologies are simulated. Experiments were demonstrated on real radio galaxy images, where we discussed the length of feature vectors, selection of lost functions, and made comparisons on batch normalization and dropout techniques for training the network. The results suggest a high efficiency and performance of our morphology generation framework. Code is available at: https://github.com/myinxd/dnnae-gmm.