Deblending galaxy superpositions with branched generative adversarial networks
Abstract
Abstract Near-future large galaxy surveys will encounter blended galaxy images at a fraction of up to 50 per cent 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 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 Large Synoptic Survey Telescope, Euclid, and Wide-Field Infrared Survey Telescope.
- Authors:
-
- Department of Physics, University of California Santa Cruz, 1156 High Street, Santa Cruz, CA 95064, USA
- Publication Date:
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1499070
- Resource Type:
- Published Article
- Journal Name:
- Monthly Notices of the Royal Astronomical Society
- Additional Journal Information:
- Journal Name: Monthly Notices of the Royal Astronomical Society Journal Volume: 485 Journal Issue: 2; Journal ID: ISSN 0035-8711
- Publisher:
- Oxford University Press
- Country of Publication:
- United Kingdom
- Language:
- English
Citation Formats
Reiman, David M., and Göhre, Brett E. Deblending galaxy superpositions with branched generative adversarial networks. United Kingdom: N. p., 2019.
Web. doi:10.1093/mnras/stz575.
Reiman, David M., & Göhre, Brett E. Deblending galaxy superpositions with branched generative adversarial networks. United Kingdom. https://doi.org/10.1093/mnras/stz575
Reiman, David M., and Göhre, Brett E. Wed .
"Deblending galaxy superpositions with branched generative adversarial networks". United Kingdom. https://doi.org/10.1093/mnras/stz575.
@article{osti_1499070,
title = {Deblending galaxy superpositions with branched generative adversarial networks},
author = {Reiman, David M. and Göhre, Brett E.},
abstractNote = {Abstract Near-future large galaxy surveys will encounter blended galaxy images at a fraction of up to 50 per cent 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 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 Large Synoptic Survey Telescope, Euclid, and Wide-Field Infrared Survey Telescope.},
doi = {10.1093/mnras/stz575},
journal = {Monthly Notices of the Royal Astronomical Society},
number = 2,
volume = 485,
place = {United Kingdom},
year = {Wed Feb 27 00:00:00 EST 2019},
month = {Wed Feb 27 00:00:00 EST 2019}
}
https://doi.org/10.1093/mnras/stz575
Web of Science
Works referenced in this record:
Galaxy Zoo: A Catalog of Overlapping Galaxy Pairs for Dust Studies
journal, January 2013
- Keel, William C.; Manning, Anna M.; Holwerda, Benne W.
- Publications of the Astronomical Society of the Pacific, Vol. 125, Issue 923
Rotation-invariant convolutional neural networks for galaxy morphology prediction
journal, April 2015
- Dieleman, Sander; Willett, Kyle W.; Dambre, Joni
- Monthly Notices of the Royal Astronomical Society, Vol. 450, Issue 2
Hyper Suprime-Cam: System design and verification of image quality
journal, September 2017
- Miyazaki, Satoshi; Komiyama, Yutaka; Kawanomoto, Satoshi
- Publications of the Astronomical Society of Japan, Vol. 70, Issue SP1
Image Quality Assessment: From Error Visibility to Structural Similarity
journal, April 2004
- Wang, Z.; Bovik, A. C.; Sheikh, H. R.
- IEEE Transactions on Image Processing, Vol. 13, Issue 4
SExtractor: Software for source extraction
journal, June 1996
- Bertin, E.; Arnouts, S.
- Astronomy and Astrophysics Supplement Series, Vol. 117, Issue 2
GalSim: The modular galaxy image simulation toolkit
journal, April 2015
- Rowe, B. T. P.; Jarvis, M.; Mandelbaum, R.
- Astronomy and Computing, Vol. 10
Wide field imaging - I. Applications of neural networks to object detection and star/galaxy classification: Wide field imaging - I
journal, December 2000
- Andreon, S.; Gargiulo, G.; Longo, G.
- Monthly Notices of the Royal Astronomical Society, Vol. 319, Issue 3
The Ellipticity Distribution of Ambiguously Blended Objects
journal, December 2015
- Dawson, William A.; Schneider, Michael D.; Tyson, J. Anthony
- The Astrophysical Journal, Vol. 816, Issue 1
scarlet: Source separation in multi-band images by Constrained Matrix Factorization
journal, July 2018
- Melchior, P.; Moolekamp, F.; Jerdee, M.
- Astronomy and Computing, Vol. 24
Multi-band morpho-Spectral Component Analysis Deblending Tool (MuSCADeT): Deblending colourful objects
journal, April 2016
- Joseph, R.; Courbin, F.; Starck, J. -L.
- Astronomy & Astrophysics, Vol. 589
The Kilo-Degree Survey
journal, August 2012
- de Jong, Jelte T. A.; Verdoes Kleijn, Gijs A.; Kuijken, Konrad H.
- Experimental Astronomy, Vol. 35, Issue 1-2
The Sixth Data Release of the Sloan Digital Sky Survey
journal, April 2008
- Adelman‐McCarthy, Jennifer K.; Agüeros, Marcel A.; Allam, Sahar S.
- The Astrophysical Journal Supplement Series, Vol. 175, Issue 2
The effective number density of galaxies for weak lensing measurements in the LSST project
journal, July 2013
- Chang, C.; Jarvis, M.; Jain, B.
- Monthly Notices of the Royal Astronomical Society, Vol. 434, Issue 3
The Dark Energy Survey: Data Release 1
journal, November 2018
- Abbott, T. M. C.; Abdalla, F. B.; Allam, S.
- The Astrophysical Journal Supplement Series, Vol. 239, Issue 2
Galaxy Zoo 1: data release of morphological classifications for nearly 900 000 galaxies★: Galaxy Zoo
journal, November 2010
- Lintott, Chris; Schawinski, Kevin; Bamford, Steven
- Monthly Notices of the Royal Astronomical Society, Vol. 410, Issue 1
Automatic analysis of crowded fields
journal, June 1985
- Irwin, M. J.
- Monthly Notices of the Royal Astronomical Society, Vol. 214, Issue 4
FOCAS - Faint Object Classification and Analysis System
journal, March 1981
- Jarvis, J. F.; Tyson, J. A.
- The Astronomical Journal, Vol. 86
Crowded Cluster Cores: An Algorithm for Deblending in Dark Energy Survey Images
journal, November 2015
- Zhang, Yuanyuan; McKay, Timothy A.; Bertin, Emmanuel
- Publications of the Astronomical Society of the Pacific, Vol. 127, Issue 957
The Morphological Evolution of Galaxies
journal, August 2001
- Abraham, R. G.
- Science, Vol. 293, Issue 5533
Photometric redshifts from SDSS images using a convolutional neural network
journal, December 2018
- Pasquet, Johanna; Bertin, E.; Treyer, M.
- Astronomy & Astrophysics, Vol. 621
Euclid mission: building of a reference survey
conference, September 2012
- Amiaux, J.; Scaramella, R.; Mellier, Y.
- SPIE Astronomical Telescopes + Instrumentation, SPIE Proceedings