From Galaxy Zoo DECaLS to BASS/MzLS: detailed galaxy morphology classification with unsupervised domain adaption - INSPIRE

DataBETA

From Galaxy Zoo DECaLS to BASS/MzLS: detailed galaxy morphology classification with unsupervised domain adaption

Renhao Ye
(
- Shanghai, Astron. Observ. and
- UCAS, Beijing
)
,
Shiyin Shen
(
- Shanghai, Astron. Observ. and
- Shanghai Normal U.
)
,
Rafael S. de Souza
(
- Hertfordshire U.
)
,
Quanfeng Xu
(
- Shanghai, Astron. Observ. and
- UCAS, Beijing
)
,
Mi Chen
(
- Shanghai, Astron. Observ. and
- UCAS, Beijing
)

Dec 19, 2024

10 pages

Published in:

Mon.Not.Roy.Astron.Soc. 537 (2025) 2, 640-649

Published: Jan 8, 2025
and
Published: Jan, 2025

e-Print:

2412.15533 [astro-ph.GA]

DOI:

10.1093/mnras/staf025

View in:

reference search0 citations

Citations per year

0 Citations

Abstract: (Oxford University Press)

The Dark Energy Spectroscopic Instrument Legacy Imaging Surveys (DESI-LIS) comprise three distinct surveys: the Dark Energy Camera Legacy Survey (DECaLS), the Beijing–Arizona Sky Survey (BASS), and the Mayall z-band Legacy Survey (MzLS). The citizen science project Galaxy Zoo DECaLS 5 (GZD-5) has provided extensive and detailed morphology labels for a sample of 253 287 galaxies within the DECaLS survey. This data set has been foundational for numerous deep learning-based galaxy morphology classification studies. However, due to differences in signal-to-noise ratios and resolutions between the DECaLS images and those from BASS and MzLS (collectively referred to as BMz), a neural network trained on DECaLS images cannot be directly applied to BMz images due to distributional mismatch. In this study, we explore an unsupervised domain adaptation (UDA) method that fine-tunes a source domain model trained on DECaLS images with GZD-5 labels to BMz images, aiming to reduce bias in galaxy morphology classification within the BMz survey. Our source domain model, used as a starting point for UDA, achieves performance on the DECaLS galaxies’ validation set comparable to the results of related works. For BMz galaxies, the fine-tuned target domain model significantly improves performance compared to the direct application of the source domain model, reaching a level comparable to that of the source domain. We also release a catalogue of detailed morphology classifications for 248 088 galaxies within the BMz survey, accompanied by usage recommendations.

Note:

11 pages, 6 figures, accepted for publication in MNRAS

methods: data analysis
galaxies: bar
galaxies: bulges
galaxies: general
galaxies: interactions

References(171)

Figures(6)

2025012316025468400_bib1

Ackermann

- Mon.Not.Roy.Astron.Soc. 479 (2018) 415
•
DOI:
- 10.1093/mnras/sty1398

Color bimodality: Implications for galaxy evolution

- AIP Conf.Proc. 743 (2004) 1, 106-119
•
e-Print:
- astro-ph/0410603
•
DOI:
- 10.1063/1.1848322

The Luminosity function of galaxies in SDSS commissioning data

SDSS

Collaboration

•

Michael R. Blanton
(
- Fermilab
)

et al.

- Astron.J. 121 (2001) 2358-2380
•
e-Print:
- astro-ph/0012085
•
DOI:
- 10.1086/320405

2025012316025468400_bib4

Ćiprijanović

2025012316025468400_bib5

Ćiprijanović

Galaxy Zoo: the fraction of merging galaxies in the SDSS and their morphologies

et al.

- Mon.Not.Roy.Astron.Soc. 401 (2010) 1043
•
e-Print:
- 0903.4937
•
DOI:
- 10.1111/j.1365-2966.2009.15686.x

Overview of the DESI Legacy Imaging Surveys

DESI

Collaboration

•

Arjun Dey
(
- NRAO, Tucson
)

et al.

- Astron.J. 157 (2019) 5, 168
•
e-Print:
- 1804.08657
•
DOI:
- 10.3847/1538-3881/ab089d

2025012316025468400_bib9

Dieleman

- Mon.Not.Roy.Astron.Soc. 450 (2015) 1441
•
DOI:
- 10.1093/mnras/stv632

2025012316025468400_bib10

Domínguez Sánchez

- Mon.Not.Roy.Astron.Soc. 476 (2018) 3661
•
DOI:
- 10.1093/mnras/sty338

Transfer learning for galaxy morphology from one survey to another

DES

Collaboration

•

H. Domínguez Sánchez
(
- Pennsylvania U.
)

et al.

- Mon.Not.Roy.Astron.Soc. 484 (2019) 1, 93-100
•
e-Print:
- 1807.00807
•
DOI:
- 10.1093/mnras/sty3497

2025012316025468400_bib12

Dullo

- Mon.Not.Roy.Astron.Soc. 462 (2016) 3800
•
DOI:
- 10.1093/mnras/stw1868

Euclid preparation. I. The Euclid Wide Survey

Euclid

Collaboration

•

R. Scaramella
(
- Rome U. and
- INFN, Rome and
- Rome Observ.
)

et al.

- Astron.Astrophys. 662 (2022) A112,
- Astron.Astrophys. 662 (2022) A112
•
e-Print:
- 2108.01201
•
DOI:
- 10.1051/0004-6361/202141938

2025012316025468400_bib6

Euclid Collaboration

2025012316025468400_bib14

Gal

2025012316025468400_bib15

Géron

- Mon.Not.Roy.Astron.Soc. 507 (2021) 4389
•
DOI:
- 10.1093/mnras/stab2064

Cosmology from the Chinese Space Station Optical Survey (CSS-OS)

et al.

- Astrophys.J. 883 (2019) 203
•
e-Print:
- 1901.04634
•
DOI:
- 10.3847/1538-4357/ab391e

2025012316025468400_bib17

González Delgado

- Astron.Astrophys. 581 (2015) A103
•
DOI:
- 10.1051/0004-6361/201525938

Explaining and Harnessing Adversarial Examples

e-Print:
- 1412.6572

The Morphological Butcher - Oemler Effect in the SDSS Cut & Enhance Galaxy Cluster Catalog

et al.

- Publ.Astron.Soc.Jap. 55 (2003) 739
•
e-Print:
- astro-ph/0301302
•
DOI:
- 10.1093/pasj/55.4.739

2025012316025468400_bib20

Hannon

•
DOI:
- 10.1093/mnras/stad2238

2025012316025468400_bib21

Hart

- Mon.Not.Roy.Astron.Soc. 461 (2016) 3663
•
DOI:
- 10.1093/mnras/stw1588

2025012316025468400_bib22

Hart

- Mon.Not.Roy.Astron.Soc. 472 (2017) 2263
•
DOI:
- 10.1093/mnras/stx2137

The Quasar Candidate Catalogs of DESI Legacy Imaging Survey Data Release 9

Zizhao He
,
Nan Li

- Res.Astron.Astrophys. 22 (2022) 9, 095021
•
e-Print:
- 2207.06792
•
DOI:
- 10.1088/1674-4527/ac839b

2025012316025468400_bib24

He

•
e-Print:
- 1911.05722

2025012316025468400_bib25

Huang

•
DOI:
- 10.1145/2046684.2046692

1-25 of 171
1
2
3
4
5
6
7
25 / page