Deblending galaxy superpositions with branched generative adversarial networks

Abstract

Near-future large galaxy surveys will encounter blended galaxy images at a fraction of up to 50% in the densest regions of the universe. Current deblending techniques may segment the foreground galaxy while leaving missing pixel intensities in the background galaxy flux. The problem is compounded by the diffuse nature of galaxies in their outer regions, making segmentation significantly more difficult than in traditional object segmentation applications. We propose a novel branched generative adversarial network (GAN) to deblend overlapping galaxies, where the two branches produce images of the two deblended galaxies. We show that generative models are a powerful engine for deblending given their innate ability to infill missing pixel values occluded by the superposition. We maintain high peak signal-to-noise ratio and structural similarity scores with respect to ground truth images upon deblending. Our model also predicts near-instantaneously, making it a natural choice for the immense quantities of data soon to be created by large surveys such as LSST, Euclid and WFIRST.

arXiv comment14 pages, 6 figures, accepted for publication in MNRAS

References in corpus (22)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Deep Residual Learning for Image Recognition
• The Sixth Data Release of the Sloan Digital Sky Survey
• NIPS 2016 Tutorial: Generative Adversarial Networks
• Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
• LSST Science Book, Version 2.0
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• Galaxy Zoo 1 : Data Release of Morphological Classifications for nearly 900,000 galaxies
• The Dark Energy Survey Data Release 1
• The Kilo-Degree Survey
• Photometric redshifts from SDSS images using a Convolutional Neural Network
• The Effective Number Density of Galaxies for Weak Lensing Measurements in the LSST Project
• SCARLET: Source separation in multi-band images by Constrained Matrix Factorization
• Wide Field Imaging. I. Applications of Neural Networks to object detection and star/galaxy classification
• The Morphological Evolution of Galaxies
• The Ellipticity Distribution of Ambiguously Blended Objects
• Euclid Mission: building of a Reference Survey
• Forging new worlds: high-resolution synthetic galaxies with chained generative adversarial networks
• Galaxy Zoo: A Catalog of Overlapping Galaxy Pairs for Dust Studies
• Multi-band morpho-Spectral Component Analysis Deblending Tool (MuSCADeT): Deblending colourful objects
• Crowded Cluster Cores: Algorithms for Deblending in Dark Energy Survey Images

Cited by in corpus (33)

• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• Deblending and Classifying Astronomical Sources with Mask R-CNN Deep Learning
• Galaxy Merger Rates up to z $\sim$ 3 using a Bayesian Deep Learning Model $-$ A Major-Merger classifier using IllustrisTNG Simulation data
• Astronomia ex machina: a history, primer, and outlook on neural networks in astronomy
• Dark Energy Survey Year 3 Results: Measuring the Survey Transfer Function with Balrog
• Photometry of high-redshift blended galaxies using deep learning
• Deep Generative Models for Galaxy Image Simulations
• Deblending galaxies with Variational Autoencoders: a joint multi-band, multi-instrument approach
• Realistic galaxy image simulation via score-based generative models
• Forging new worlds: high-resolution synthetic galaxies with chained generative adversarial networks
• Painting halos from cosmic density fields of dark matter with physically motivated neural networks
• Differentiable Strong Lensing: Uniting Gravity and Neural Nets through Differentiable Probabilistic Programming
• Optimising and comparing source extraction tools using objective segmentation quality criteria
• Detection, Instance Segmentation, and Classification for Astronomical Surveys with Deep Learning (DeepDISC): Detectron2 Implementation and Demonstration with Hyper Suprime-Cam Data
• 21st Century Statistical and Computational Challenges in Astrophysics
• Deblending Galaxies with Generative Adversarial Networks
• Gaussian Process Classification for Galaxy Blend Identification in LSST
• Galaxy Deblending using Residual Dense Neural networks
• Galaxy Image Translation with Semi-supervised Noise-reconstructed Generative Adversarial Networks
• Machine learning technique for morphological classification of galaxies from SDSS. II. The image-based morphological catalogs of galaxies at 0.02<z<0.1
• Machine learning technique for morphological classification of galaxies from the SDSS. III. Image-based inference of detailed features
• Deep Transfer Learning for Blended Source Identification in Galaxy Survey Data
• Variational Inference for Deblending Crowded Starfields
• Hybrid Physical-Deep Learning Model for Astronomical Inverse Problems
• Score-matching neural networks for improved multi-band source separation
• Deblending Overlapping Galaxies in DECaLS Using Transformer-Based Algorithm: A Method Combining Multiple Bands and Data Types
• Using machine learning to study the kinematics of cold gas in galaxies
• Blending effects on shear measurement synergy between Euclid-like and LSST-like surveys
• Investigation on deep learning-based galaxy image translation models
• Astro2020 Science White Paper: The Next Decade of Astroinformatics and Astrostatistics
• A Simulation Driven Deep Learning Approach for Separating Mergers and Star Forming Galaxies: The Formation Histories of Clumpy Galaxies in all the CANDELS Fields
• Interpreting Galaxy Deblender GAN from the Discriminator's Perspective