CoChat
Sign Up
Cited by 28Galaxies: Formation, Evolution, PhenomenaAdvanced Neural Network ApplicationsAstrophysical Phenomena and Observations

Identification of tidal features in deep optical galaxy images with convolutional neural networks

Open access
Sánchez, H Domínguez, Martin, G, Damjanov, I, Buitrago, F, Huertas-Company, M, Bottrell, C, Bernardi, M, Knapen, J H +9 more·2023·Monthly Notices of the Royal Astronomical Society
DOI 10.1093/mnras/stad750
Open this paper in CoChat — free accountOpen PDF

Abstract

Interactions between galaxies leave distinguishable imprints in the form of tidal features, which hold important clues about their mass assembly. Unfortunately, these structures are difficult to detect because they are low surface brightness features, so deep observations are needed. Upcoming surveys promise several orders of magnitude increase in depth and sky coverage, for which automated methods for tidal feature detection will become mandatory. We test the ability of a convolutional neural network to reproduce human visual classifications for tidal detections. We use as training ∼6000 simulated images classified by professional astronomers. The mock Hyper Suprime Cam Subaru (HSC) images include variations with redshift, projection angle, and surface brightness (μlim = 26–35 mag arcsec−2). We obtain satisfactory results with accuracy, precision, and recall values of Acc = 0.84, P = 0.72, and R = 0.85 for the test sample. While the accuracy and precision values are roughly constant for all surface brightness, the recall (completeness) is significantly affected by image depth. The recovery rate shows strong dependence on the type of tidal features: we recover all the images showing shell features and 87 per cent of the tidal streams; these fractions are below 75 per cent for mergers, tidal tails, and bridges. When applied to real HSC images, the performance of the model worsens significantly. We speculate that this is due to the lack of realism of the simulations, and take it as a warning on applying deep learning models to different data domains without prior testing on the actual data.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/mnras/stad750

Cites

91 of 91 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Domínguez Sánchez et al. (2023) found at surface brightness limits close to HSC-SSP UD limits, model reached ROC AUC = 0.93 and completeness ~85% for 22% contamination.

    “They also separated the performance of the model according to the image surface brightness limits and found that at surface brightness limits close to HSC-SSP UD limits ( 𝜇 𝑟 ∼ 30magarcsec -2 ) the model reached ROC AUC = 0.93 and completeness of ∼ 85% for 22% contamination.”

    From Detecting galaxy tidal features using self-supervised representation learning (2024) · cites this paper as Domínguez Sánchez et al. (2023)

    Section 4.3 (page not available in our copy) reports ROC AUC = 0.93 at a surface-brightness limit of 30 mag arcsec⁻². At that limit, however, it reports 90% of tidal detections recovered and 26% contamination, not approximately 85% and 22%. The reported performance is for mock images; the cited passages do not establish that this depth is close to the HSC-SSP UD limit. (LLM, conf=0.91)

    Experimental — accuracy not yet measured

  • The ROC AUC reached by our model (0.911 ± 0.002) is comparable to that found in Domínguez Sánchez et al. (2023), however, we reach a significantly higher completeness of 96% for the same level of contamination.

    “The ROC AUC reached by our model (0.911 ± 0.002) is comparable to that found in Domínguez Sánchez et al. (2023), however, we reach a significantly higher completeness of 96% for the same level of contamination.”

    From Detecting galaxy tidal features using self-supervised representation learning (2024) · cites this paper as Domínguez Sánchez et al. (2023)

  • Domínguez Sánchez et al. (2023) found applying their trained model to real HSC-SSP images decreased performance to ROC AUC = 0.64.

    “Domínguez Sánchez et al. (2023) also attempted to apply their trained model to real HSC-SSP images, however the model performance decreased significantly, only reaching ROC AUC = 0.64.”

    From Detecting galaxy tidal features using self-supervised representation learning (2024) · cites this paper as Domínguez Sánchez et al. (2023)

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 8

  1. [11]11. Chen Z., Fu Y., Zhang Y., Jiang Y.-G., Xue X., Sigal L., 2019, IEEE Trans- actions on Image Processing, 28, 4594
  2. [17]17. Conselice C. J., Mundy C. J., Ferreira L., Duncan K., 2022, arXiv e-prints, p. arXiv:2207.03984 · arXiv
  3. [32]32. He K., Zhang X., Ren S., Sun J., 2016, in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp 770-778, doi:10.1109/CVPR.2016. · DOI
  4. [38]38. Huertas-Company M., Lanusse F., 2022, arXiv e-prints, p. arXiv:2210.01813 · arXiv
  5. [41]41. Huško F., Lacey C. G., Baugh C. M., 2022, arXiv e-prints, p. arXiv:2207.07139 · arXiv
  6. [64]64. Martinez-Delgado D., et al., 2021, arXiv e-prints, p. arXiv:2104.06071 · arXiv
  7. [72]72. Pearson S., Price-Whelan A. M., Hogg D. W., Seth A. C., Sand D. J., Hunt J. A. S., Crnojević D., 2022, ApJ, 941, 19
  8. [86]86. Valenzuela L. M., Remus R.-S., 2022, arXiv e-prints, p. arXiv:2208.08443 · arXiv

Cited by

Sorted by citations

Related

Authors
Sánchez, H. D., Martin, G., Damjanov, I., Buitrago, F., Huertas-Company, M., Bottrell, C., Bernardi, M., Knapen, J. H., Vega-Ferrero, J., Hausen, R., Kado-Fong, E., Población-Criado, D., Souchereau, H., Leste, O. K., Robertson, B., Sahelices, B., & Johnston, K. V.
Affiliations
  • Institute of Space Sciences (ES)
  • University of Arizona (US)
  • Saint Mary's University (CA)
  • Universidad de Valladolid (ES)
  • Instituto de Astrofísica de Canarias (IAC) , La Laguna E-38205, Spain
  • Universidad de La Laguna (ES)
  • Kavli Institute for the Physics and Mathematics of the Universe (JP)
  • Johns Hopkins University (US)
  • University of Victoria (CA)
  • University of California, Santa Cruz (US)
Corresponding author
Sánchez, H Domínguez
ORCID
Venue
Monthly Notices of the Royal Astronomical Society
Publisher
Oxford University Press (OUP)
Published
2023-03-22
Volume
521
Issue
3
Pages
3861-3872
ISSN
  • 0035-8711 (ISSN-L)
  • 1365-2966
DOI
10.1093/mnras/stad750
OpenAlex
W4353029737
Cited by
28
References
89
Citations by year
2026: 7 · 2025: 7 · 2024: 11 · 2023: 3
FWCI
3.26
Type
journal article
Language
en
Open access
Licence
academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model (vor)
Locations
  • Monthly Notices of the Royal Astronomical Society (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • UA Campus Repository (The University of Arizona) (submittedVersion)
  • DIGITAL.CSIC (Spanish National Research Council (CSIC)) (submittedVersion)
  • Liverpool John Moores University (submittedVersion)
  • arXiv (Cornell University)
Funders
  • AEI (10.13039/501100011033)
  • Spanish Ministry of Science and Innovation (PGC2018-100852-A-I00, PID2019-105602GB-I00/10.13039/501100011033)
  • European Union
  • Natural Sciences and Engineering Research Council of Canada (RGPIN-2018-05425, PID2020-116188GA-I00 and PID2019-107427GB-C32)
  • Spanish State Research Agency
  • ACIISI
  • European Regional Development Fund (PROID2021010044)
  • IAC (P/300724)
Topics
  • Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Galaxies: Formation, Evolution, Phenomena
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Astrophysical Phenomena and Observations
Keywords
tidal features, galaxy interactions, convolutional neural networks, automated feature detection, low surface brightness, tidal currents, tidal tails, galaxy mergers, tidal bridges, galaxy mass assembly, domain generalization

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review