Pith. sign in

REVIEW 4 major objections 5 minor 99 references

Classification of Spiral Galaxies by Spiral Arm Number using Convolutional Neural Network

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Two convolutional classifiers can sort spiral galaxies by arm number with most F1 scores above 0.8, except for the rare four-arm class; merging it with the five-plus class lifts the V2M model's recall to about 0.84.

desk verdict A useful GZ2 arm-count classifier with credible results for common classes, but rare-class metrics are statistically thin, the B0 model inherits the target question from Zoobot, and the abstract overstates the m=3 stellar mass result. read the letter →

arxiv 2412.11696 v1 pith:SVJLRHDG submitted 2024-12-16 astro-ph.GA

Ming Wei Lee , John Y. H. Soo , Syarawi M. H. Sharoni This is my paper · ORCID
classification astro-ph.GA
keywords spiralarmnumbergalaxymorphologyclassificationconvolutionalneuralnetworkEfficientNetZoo2transferlearningstellarmassdeep
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a straightforward CNN pipeline—no arm-segmentation or sophisticated structure analysis—can classify spiral galaxies by the number of spiral arms accurately enough for large surveys. Using Galaxy Zoo 2 vote-fraction labels filtered to a clean sample of 11,718 images, both an ImageNet-fine-tuned EfficientNetV2M and a Zoobot-feature EfficientNetB0 reach F1 scores above 0.8 for most arm-number classes on a balanced down-sampled set. The four-arm class is the persistent failure, with only 28 training images and recall near random (0.222). Merging four-arm and five-plus-arm galaxies repairs the V2M model's performance for that combined class, at the cost of the scientifically meaningful four-arm distinction. The authors further connect arm number to stellar mass, finding m=1 galaxies skew massive and m=3 galaxies skew light, consistent with earlier Galaxy Zoo work.

What carries the argument

The load-bearing machinery is transfer learning on EfficientNet backbones: a fine-tuned EfficientNetV2M initialized with ImageNet weights, and an EfficientNetB0 whose weights are frozen from the Zoobot galaxy-morphology model, each topped with a global average pooling layer, a dense ReLU layer, and a six-class softmax. The training set is the crucial second piece: Galaxy Zoo 2 images pre-filtered by weighted and debiased vote-fraction cuts (rho_features/disc>0.430, rho_not edge-on>0.715, rho_spiral,yes>0.715, N_spiral,yes>20, rho_m>0.8), down-sampled to roughly 300 per class with rotation, flipping, and zoom augmentation. Gradient heatmaps (GradCAM++) and saliency maps (SmoothGrad) carry the interpretability argument, showing that decisions track spiral structure rather than background artifacts for correct predictions.

What would settle it

Re-train both models on a larger, independent sample with expert-confirmed arm counts, or on many random 70:30 splits of the same data, and check whether m=4 recall stays near 0.222 and whether the merged-class improvement holds; a stable m=4 recall well above 0.5 on independent labels would refute the sample-size explanation, while a collapse across splits would show the headline metrics are artifacts of split composition.

Watch

Extended reading notes

Core claim

The central claim is that spiral arm number is learnable from raw survey images by generic CNN transfer learning, and that classifier labels reproduce known physical trends well enough to be usable at survey scale. On a down-sampled Galaxy Zoo 2 dataset with about 300 images per class, the V2M model (EfficientNetV2M fine-tuned on ImageNet) and the B0 model (EfficientNetB0 with frozen Zoobot weights) both deliver precision, recall, and F1 above 0.8 for m=1, 2, 3, and 'can't tell', and near or above 0.8 for m=5+, with m=4 the only class at chance-level recall. Combining m=4 with m=5+ raises the V2M model's recall for the merged class to about 0.84, which the authors argue balances scientific discrimination against the scarcity of high-arm samples. GradCAM++ and SmoothGrad indicate the networks base decisions on galaxy structure and spiral-arm extraction, with V2M the stronger of the two. On physical tests, the GZ2 labels show a significant shift of m=1 galaxies toward higher stellar mass and a weaker shift of m=3 toward lower mass; both tendencies are attenuated in model predictions.

Load-bearing premise

The whole evaluation rests on the Galaxy Zoo 2 debiased vote-fraction labels being correct for every class, especially the 28 four-arm and 31 five-plus-arm galaxies; if those rare labels are noisy or the selection cuts exclude certain galaxy types, the reported F1 scores and the stellar-mass trends collapse.

Editorial extensions

If this is right

  • If reliable, these classifiers give a simple, scalable way to assign spiral arm numbers to the millions of galaxies expected from Euclid, Roman, and Rubin/LSST without human inspection.
  • Merging m=4 and m=5+ is a practical class scheme that preserves the low/high-arm distinction while keeping V2M F1 above 0.8 for every class.
  • The attenuated stellar-mass trends in model predictions imply that classifier noise dilutes physical correlations, so any survey-scale arm-number catalog should budget for that dilution.
  • The m=4 failure quantifies the sample-size floor: with about 28 clean examples, even strong transfer learning cannot learn the class, suggesting dedicated collection or synthesis is needed.
  • The alignment of predictions with the known m=1 high-mass and m=3 low-mass tendencies, despite dilution, suggests the labels carry real physical signal worth pursuing.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The tiny m=4 class (28 images total, about 8–10 in validation) makes the reported 0.222 recall extremely sensitive to which images land in the validation split; refitting across multiple random splits would show whether the m=4 numbers are a stable property or split luck.
  • A testable extension is to run the same frozen-backbone B0 pipeline on deeper imaging (e.g., Stripe-82 coadds or DECaLS) where faint high-arm structure is more visible; if m=4 and m=5+ recall improves markedly, the current limitation is image depth, not class distinctness.
  • The results suggest a practical rule for other morphology tasks: when a rare class sits at the edge of a decision tree, merging it with its nearest neighbor and reporting both merged and unmerged metrics is more informative than forcing a six-way split.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This paper trains two CNN classifiers (EfficientNetV2M fine-tuned on ImageNet, and EfficientNetB0 with frozen Zoobot weights) to classify galaxy images from Galaxy Zoo 2 into spiral arm numbers m=1, 2, 3, 4, 5+ and 'can't tell'. The dataset is filtered with GZ2 vote-fraction cuts (Section 2.1) and down-sampled to balance classes. The authors report precision/recall/F1 in Table 3, class-combination experiments with confusion matrices (Figure 7), Grad-CAM++ and SmoothGrad visualizations, and a comparison of stellar mass distributions between GZ2 labels and model predictions (Section 6). They conclude that both models achieve high accuracy except for m=4, and that merging m=4 and m=5+ improves V2M performance.

Significance. The paper's strengths are its detailed and conservative dataset selection, transparent reporting of confusion matrices, and direct comparison of two transfer-learning strategies. If the performance numbers are robust, the classifiers would be a useful, scalable tool for arm-number counting in upcoming surveys. However, the central accuracy claims for the rare classes m=4 and m=5+ are statistically fragile because they rest on roughly 8-19 validation images without uncertainty quantification, and the B0 comparison is complicated by the Zoobot backbone having been pre-trained on the same GZ decision-tree question. These issues are fixable but currently limit the strength of the conclusions.

major comments (4)
  1. [Section 5.1, Table 3] The headline metrics for the two rarest classes are point estimates from a single 70:30 split with no uncertainty quantification. With 28 m=4 and 31 m=5+ images in the full dataset, the validation sets contain roughly 8-10 images; the reported m=4 recall of 0.222 therefore corresponds to about 2 correct of about 9 images, so a single image changes recall by ~0.11 and F1 by ~0.15. The m=5+ precision of 0.769 for V2M is also below the stated 0.8 baseline. Please report per-class validation counts and either Wilson/binomial confidence intervals or repeated-split / k-fold distributions for all metrics; without this, the abstract's 'except m=4' clause rests on a statistically fragile class.
  2. [Section 5.2, Figure 7] The class-combination result is likewise based on about 19 validation images for the merged m=4+5+ class. V2M recall 0.842 corresponds to 16/19 correct, and the 95% Wilson interval is approximately (0.60, 0.95); the '280% improvement' quoted later is therefore not a precisely determined quantity. Please report raw counts for the merged class and include confidence intervals, and state explicitly that the improvement is in relative recall rather than accuracy.
  3. [Section 3.3, Section 5.1] The B0 model uses a frozen Zoobot backbone that was pre-trained to predict GZ decision tree responses, including the arm-count question (T11/Q11), so its high F1 on that same question may largely reflect transfer from pre-training rather than a newly learned classifier. This is a correctness risk for the B0 vs. V2M comparison, not an assertion of intentional leakage. A concrete control would be to compare B0 against an unfrozen EfficientNetB0, to report how the raw Zoobot model performs on this same validation split, or to quantify the overlap between the DECaLS training set used by Zoobot and the GZ2 main sample used here.
  4. [Section 6] The stellar-mass analysis uses model predictions on 'our entire data set (including all training, validating and testing images)', so the predicted-class distributions include in-sample predictions on training data. This can bias the comparison between label and prediction distributions and weakens the claim that the mass tendencies are 'reduced in the model predictions'. Please recompute the predicted distributions using only the validation (or a nested) split, or justify why training-set contamination does not affect the conclusion.
minor comments (5)
  1. [Abstract] There is a grammar issue: 'the spiral arm number, offer' should read 'the spiral arm number offers'.
  2. [Table 2] Typo: 'Number of hyparameters' should be 'Number of hyperparameters'.
  3. [Section 4.2.1] The reported coefficient of variation values 2.203 and 2.124 are inconsistent with F1-scores near 0.9 if CV is defined as sigma/mu in the standard sense; please check whether these are percentages or have a missing factor of 100, and define the convention explicitly.
  4. [Section 5.2 text and Figure 7 caption] The text says 'A considerably high percentage of m = 3 galaxies are misclassified as m = 4 (67%)', but the confusion matrices show the opposite: 66.7% of m=4 galaxies are misclassified as m=3. The Figure 7 caption's 'Up to 22% of m = 3 galaxies are misclassified as m = 4' also does not match the shown values of 13.3% (B0) and 6.7% (V2M).
  5. [Section 7] The '280% accuracy improvement' for the V2M merged class should be stated as a relative improvement in recall (from 0.222 to 0.842), not as an improvement in accuracy.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper reports empirical CNN performance against external GZ2 labels; no derivation or fitted law reduces to its inputs.

full rationale

This paper does not contain a derivation chain of the kind that can be circular: its central claims are measured precision, recall, and F1 values from supervised CNNs trained on GZ2 images and evaluated on a 70:30 holdout (Sections 3.5 and 5.1, Table 3). The ground-truth labels are external Galaxy Zoo 2 volunteer vote aggregates (Section 2.1), not values computed from the model. The image-size and class-balancing optimizations (Sections 4.2.1 and 4.2.2) are model-selection steps, not fitted parameters renamed as predictions. The B0 model uses a frozen Zoobot backbone that was pretrained on DECaLS imaging to predict GZ decision-tree responses including the arm-count question (Section 3.3); this is a domain-specific transfer-learning choice and a legitimate independence concern for the B0 comparison, but it is not circular by construction: the classification head is retrained on the downsampled GZ2 set, and the reported metrics are computed on held-out GZ2 images that were not used to train that head. The paper explicitly notes the limited samples in the m=4 and m=5+ classes and the need for further improvements for high-arm galaxies (Sections 5.1 and 7), which is a statistical and data limitation rather than circularity. No self-citations are load-bearing, no uniqueness theorem is imported from the authors, and no known result is merely renamed.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities. The central claims rest on hand-chosen data filtering and preprocessing choices (a 224x224 crop, a 300-image-per-class cap) and on assumptions about label quality and catalog reliability.

free parameters (2)
  • Crop size = 224
    Chosen as optimal in a validation sweep over image sizes (Section 4.2.1), balancing F1 and coefficient of variation; a hand-tuned hyperparameter that affects all reported results.
  • Downsampling cap = 300
    Selected to balance the class distribution (Section 4.2.2); the cap of 300 images per class determines the training and validation sets used for the main results.
assumptions (4)
  • domain assumption GZ2 volunteer labels, aggregated with weighted and debiased vote fractions, are accurate enough to serve as ground truth
    The entire training and evaluation rests on the rho_m>0.8 threshold and vote-count cuts from Willett et al. (2013), described in Section 2.1. Noisy or biased labels would directly inflate or distort the reported accuracies and the stellar mass trends.
  • domain assumption The MPA-JHU stellar mass estimates (LGM_TOT_P50, Kroupa IMF) are reliable
    Section 6 uses these catalog values for the physical correlation analysis; their uncertainties are not propagated into the K-S or t-test results.
  • domain assumption Zoobot pre-training on GZ decision tree responses is a valid feature extractor for this task
    Section 3.3; the B0 model's performance depends on this premise. Because Zoobot was trained on the same arm-count question, this premise also introduces potential label leakage that is not discussed.
  • domain assumption The random 70:30 split of the downsampled dataset is representative, particularly for classes with fewer than 35 images
    Section 3.5; with m=4 at 28 images and m=5+ at 31 images, the validation subsets are roughly 8-10 images each, so the realized metrics depend heavily on the random split composition.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Classification of Spiral Galaxies by Spiral Arm Number using Convolutional Neural Network." pith.science (2026). https://pith.science/paper/SVJLRHDG

@misc{pith2026241211696,
  author       = {Pith},
  title        = {Pith review of: Classification of Spiral Galaxies by Spiral Arm Number using Convolutional Neural Network},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SVJLRHDG}},
  note         = {Machine review of arXiv:2412.11696}
}
read the original abstract

The structural information of spiral galaxies such as the spiral arm number, offer valuable insights into their formation processes and physical roles in galaxy evolution. We developed classifiers based on CNNs using variants of the EfficientNet architecture with different transfer learning techniques and pre-trained weights to categorise spiral galaxies by their number of spiral arms. A dataset from GZ2, comprising 11718 images filtered based on appropriate criteria is used for training and evaluation. Both the EfficientNetV2M model fine-tuned on ImageNet and the EfficientNetB0 model with Zoobot pre-trained weights achieved high accuracy on the down-sampled dataset, with most performance metrics exceeding 0.8 across all classes, except for galaxies with 4 arms due to the limited number of samples in this category. Merging higher-arm-number classes (more than 4 arms) improved the V2M model's accuracy significantly for 4-arm galaxies, as this approach allowed the model to focus on more distinct features with a more balanced class distribution. GradCAM++ and SmoothGrad highlight the networks' effectiveness in classifying galaxies, through the distinction of the galaxy structures and the extraction of the spiral arms, with the V2M model showing better capabilities in both tasks. Lower-arm galaxies tend to be misclassified as "can't tell" when their spiral arms are not clearly visible, while higher-arm galaxies tend to be misclassified as having fewer arms when their features are only partially detected. The study also found that galaxies with 3 arms tend to have lower stellar masses, and this tendency is reduced in the model predictions. The models' mispredictions between 2-arm and 1/3-arm are likely resulting from external interference and dynamic nature of spiral arms. The V2M model prediction also shows a slight tendency towards higher stellar mass in high-arm galaxies.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

99 extracted references · 73 canonical work pages

  1. [1]

    Hubble, Extragalactic nebulae., ApJ 64 (1926) 321

    E.P. Hubble, Extragalactic nebulae., ApJ 64 (1926) 321

    1926

  2. [2]

    Kennicutt, Star formation in galaxies along the Hubble sequence , Annual Review of A&A 36 (1998) 189

    R.C.J. Kennicutt, Star formation in galaxies along the Hubble sequence , Annual Review of A&A 36 (1998) 189

    1998

  3. [3]

    Masters, C.J

    K.L. Masters, C.J. Lintott, R.E. Hart, S.J. Kruk, R.J. Smethurst, K.V. Casteels et al., Galaxy Zoo: unwinding the winding problem–observations of spiral bulge prominence and arm pitch angles suggest local spiral galaxies are winding , MNRAS 487 (2019) 1808

    2019

  4. [4]

    De Vaucouleurs, Classification and morphology of external galaxies , Handbuch der Physik 53 (1959) 275

    G. De Vaucouleurs, Classification and morphology of external galaxies , Handbuch der Physik 53 (1959) 275

    1959

  5. [5]

    Elmegreen and B.G

    D.M. Elmegreen and B.G. Elmegreen, Flocculent and grand design spiral structure in field, binary and group galaxies , MNRAS 201 (1982) 1021

    1982

  6. [6]

    Elmegreen and B.G

    D.M. Elmegreen and B.G. Elmegreen, Arm classifications for spiral galaxies , ApJ 314 (1987) 3

    1987

  7. [7]

    Lintott, K

    C. Lintott, K. Schawinski, S. Bamford, A. Slosar, K. Land, D. Thomas et al., Galaxy Zoo 1: data release of morphological classifications for nearly 900000 galaxies , MNRAS 410 (2011) 166

    2011

  8. [8]

    Willett, C.J

    K.W. Willett, C.J. Lintott, S.P. Bamford, K.L. Masters, B.D. Simmons, K.R.V. Casteels et al., Galaxy Zoo 2: detailed morphological classifications for 304122 galaxies from the Sloan Digital Sky Survey , MNRAS 435 (2013) 2835

    2013

Show all 99 references
  1. [9]

    Lindblad and R.G

    B. Lindblad and R.G. Langebartel, On the dynamics of stellar systems , Stockholms Observatoriums Annaler 17 (1953) 6

    1953

  2. [10]

    Lin and F.H

    C.C. Lin and F.H. Shu, On the spiral structure of disk galaxies , ApJ 140 (1964) 646

    1964

  3. [11]

    Goldreich and D

    P. Goldreich and D. Lynden-Bell, Ii. spiral arms as sheared gravitational instabilities , MNRAS 130 (1965) 125

    1965

  4. [12]

    Julian and A

    W.H. Julian and A. Toomre, Non-axisymmetric responses of differentially rotating disks of stars , ApJ 146 (1966) 810

    1966

  5. [13]

    Toomre and J

    A. Toomre and J. Toomre, Galactic bridges and tails , ApJ 178 (1972) 623

    1972

  6. [14]

    Hubble, The direction of rotation in spiral nebulae

    E.P. Hubble, The direction of rotation in spiral nebulae. , ApJ 97 (1943) 112

    1943

  7. [15]

    Elmegreen and M

    B.G. Elmegreen and M. Thomasson, Grand design and flocculent spiral structure in computer simulations with star formation and gas heating , A&A 272 (1993) 37

    1993

  8. [16]

    Zhang, Secular evolution of spiral galaxies

    X. Zhang, Secular evolution of spiral galaxies. I. a collective dissipation process , ApJ 457 (1996) 125

    1996

  9. [17]

    Block and R.J

    D.L. Block and R.J. Wainscoatt, Morphological differences between optical and infrared images of the spiral galaxy NGC309 , Nature 353 (1991) 48

    1991

  10. [18]

    Block, G

    D.L. Block, G. Bertin, A. Stockton, P. Grosbøl, A.F.M. Moorwood and R.F. Peletier, 2.1 µm images of the evolved stellar disk and the morphological classification of spiral galaxies , A&A 288 (1994) 365. – 24 –

    1994

  11. [19]

    Shabani, E.K

    F. Shabani, E.K. Grebel, A. Pasquali, E. D ´Onghia, J.S. Gallagher III, A. Adamo et al., Search for star cluster age gradients across spiral arms of three LEGUS disc galaxies , MNRAS 478 (2018) 3590

    2018

  12. [20]

    Peterken, M.R

    T.G. Peterken, M.R. Merrifield, A. Arag´ on-Salamanca, N. Drory, C.M. Krawczyk, K.L. Masters et al., A direct test of density wave theory in a grand-design spiral galaxy , Nature Astronomy 3 (2019) 178

    2019

  13. [21]

    Bialopetraviˇ cius and D

    J. Bialopetraviˇ cius and D. Narbutis,Study of star clusters in the M83 galaxy with a convolutional neural network , AJ 160 (2020) 264

    2020

  14. [22]

    Abdeen, B.L

    S. Abdeen, B.L. Davis, R. Eufrasio, D. Kennefick, J. Kennefick, R. Miller et al., Evidence in favour of density wave theory through age gradients observed in star formation history maps and spatially resolved stellar clusters , MNRAS 512 (2022) 366

    2022

  15. [23]

    Kormendy and C.A

    J. Kormendy and C.A. Norman, Observational constraints on driving mechanisms for spiral density waves , ApJ 233 (1979) 539

    1979

  16. [24]

    Grosbøl, P.A

    P. Grosbøl, P.A. Patsis and E. Pompei, Spiral galaxies observed in the near-infrared K band–I. Data analysis and structural parameters , A&A 423 (2004) 849

    2004

  17. [25]

    Grosbøl and G

    P. Grosbøl and G. Carraro, Is the Galactic Spiral Potential 2-or 4-arms? , Proc. IAU 334 (2018) 300

    2018

  18. [26]

    Khrapov, A

    S. Khrapov, A. Khoperskov and V. Korchagin, Modeling of spiral structure in a multi-component Milky Way-like galaxy , Galaxies 9 (2021) 29

    2021

  19. [27]

    Hart, S.P

    R.E. Hart, S.P. Bamford, W.C. Keel, S.J. Kruk, K.L. Masters, B.D. Simmons et al., Galaxy Zoo: constraining the origin of spiral arms , MNRAS 478 (2018) 932

    2018

  20. [28]

    Hart, S.P

    R.E. Hart, S.P. Bamford, K.W. Willett, K.L. Masters, C. Cardamone, C.J. Lintott et al., Galaxy Zoo: comparing the demographics of spiral arm number and a new method for correcting redshift bias, MNRAS 461 (2016) 3663

    2016

  21. [29]

    van den Bergh, A new classification system for galaxies , ApJ 206 (1976) 883

    S. van den Bergh, A new classification system for galaxies , ApJ 206 (1976) 883

    1976

  22. [30]

    Abraham, F

    R.G. Abraham, F. Valdes, H.K.C. Yee and S. van den Bergh, The morphologies of distant galaxies. I. an automated classification system , ApJ 432 (1994) 75

    1994

  23. [31]

    Abraham, N.R

    R.G. Abraham, N.R. Tanvir, B.X. Santiago, R.S. Ellis, K. Glazebrook and S. van den Bergh, Galaxy morphology to I=25 mag in the Hubble Deep Field , MNRAS 279 (1996) L47

    1996

  24. [32]

    Abraham, S

    R.G. Abraham, S. van den Bergh, K. Glazebrook, R.S. Ellis, B.X. Santiago, P. Surma et al., The morphologies of distant galaxies. II. Classifications from the Hubble Space Telescope Medium Deep Survey, ApJSS 107 (1996) 1

    1996

  25. [33]

    Conselice, M.A

    C.J. Conselice, M.A. Bershady and A. Jangren, The asymmetry of galaxies: Physical morphology for nearby and high-redshift galaxies , ApJ 529 (2000) 886

    2000

  26. [34]

    A. Naim, O. Lahav, L.J. Sodre and M.C. Storrie-Lombardi, Automated morphological classification of APM galaxies by supervised artificial neural networks , MNRAS 275 (1995) 567

    1995

  27. [35]

    Lahav, A

    O. Lahav, A. Naim, L.J. Sodr´ e and M.C. Storrie-Lombardi, Neural computation as a tool for galaxy classification: methods and examples , MNRAS 283 (1996) 207

    1996

  28. [36]

    Owens, R.E

    E.A. Owens, R.E. Griffiths and K.U. Ratnatunga, Using oblique decision trees for the morphological classification of galaxies , MNRAS 281 (1996) 153

    1996

  29. [37]

    Bazell and D.W

    D. Bazell and D.W. Aha, Ensembles of classifiers for morphological galaxy classification , ApJ 548 (2001) 219

    2001

  30. [38]

    Goderya and S.M

    S.N. Goderya and S.M. Lolling, Morphological classification of galaxies using computer vision and artificial neural networks: A computational scheme , A&SS 279 (2002) 377. – 25 –

    2002

  31. [39]

    N.M. Ball, J. Loveday, M. Fukugita, O. Nakamura, S. Okamura, J. Brinkmann et al., Galaxy types in the Sloan Digital Sky Survey using supervised artificial neural networks , MNRAS 348 (2004) 1038

    2004

  32. [40]

    Huertas-Company, D

    M. Huertas-Company, D. Rouan, L. Tasca, G. Soucail and O. Le F` evre, A robust morphological classification of high-redshift galaxies using support vector machines on seeing limited images. I. method description, A&A 478 (2008) 971

    2008

  33. [41]

    Huertas-Company, L

    M. Huertas-Company, L. Tasca, D. Rouan, D. Pelat, J.P. Kneib, O. Le F` evre et al., A robust morphological classification of high-redshift galaxies using support vector machines on seeing limited images. II. Quantifying morphological k-correction in the COSMOS field at 1 < z <...

    2009

  34. [42]

    Huertas-Company, J.A.L

    M. Huertas-Company, J.A.L. Aguerri, M. Bernardi, S. Mei and J. S´ anchez Almeida, Revisiting the Hubble sequence in the SDSS DR7 spectroscopic sample: a publicly available Bayesian automated classification, A&A 525 (2011) A157

    2011

  35. [43]

    Banerji, O

    M. Banerji, O. Lahav, C.J. Lintott, F.B. Abdalla, K. Schawinski, S.P. Bamford et al., Galaxy Zoo: reproducing galaxy morphologies via machine learning , MNRAS 406 (2010) 342

    2010

  36. [44]

    D.G. York, J. Adelman, J.E. Anderson Jr, S.F. Anderson, J. Annis, N.A. Bahcall et al., The sloan digital sky survey: Technical summary , AJ 120 (2000) 1579

    2000

  37. [45]

    Cheng, C.J

    T.-Y. Cheng, C.J. Conselice, A. Arag´ on-Salamanca, N. Li, A.F.L. Bluck, W.G. Hartley et al., Optimizing automatic morphological classification of galaxies with machine learning and deep learning using Dark Energy Survey imaging , MNRAS 493 (2020) 4209

    2020

  38. [46]

    Dieleman, K.W

    S. Dieleman, K.W. Willett and J. Dambre, Rotation-invariant convolutional neural networks for galaxy morphology prediction, MNRAS 450 (2015) 1441

    2015

  39. [47]

    Dom ´ ınguez S´ anchez, M

    H. Dom ´ ınguez S´ anchez, M. Huertas-Company, M. Bernardi, D. Tuccillo and J.L. Fischer, Improving galaxy morphologies for SDSS with deep learning , MNRAS 476 (2018) 3661

    2018

  40. [48]

    Zhu, J.-M

    X.-P. Zhu, J.-M. Dai, C.-J. Bian, Y. Chen, S. Chen and C. Hu, Galaxy morphology classification with deep convolutional neural networks , A&SS 364 (2019) 55

    2019

  41. [49]

    Kalvankar, H

    S. Kalvankar, H. Pandit and P. Parwate, Galaxy morphology classification using EfficientNet architectures, ArXiv e-prints (2020) [ 2008.13611]

    2020 arXiv

  42. [50]

    Cavanagh, K

    M.K. Cavanagh, K. Bekki and B.A. Groves, Morphological classification of galaxies with deep learning: comparing 3-way and 4-way CNNs , MNRAS 506 (2021) 659

    2021

  43. [51]

    Hart, S.P

    R.E. Hart, S.P. Bamford, W.B. Hayes, C.N. Cardamone, W.C. Keel, S.J. Kruk et al., Galaxy Zoo and SPARCFIRE: constraints on spiral arm formation mechanisms from spiral arm number and pitch angles , MNRAS 472 (2017) 2263

    2017

  44. [52]

    Davis and W.B

    D.R. Davis and W.B. Hayes, SpArcFiRe: scalable automated detection of spiral galaxy arm segments, ApJ 790 (2014) 87

    2014

  45. [53]

    Bekki, Quantifying the fine structures of disk galaxies with deep learning: Segmentation of spiral arms in different Hubble types , A&A 647 (2021) A120

    K. Bekki, Quantifying the fine structures of disk galaxies with deep learning: Segmentation of spiral arms in different Hubble types , A&A 647 (2021) A120

    2021

  46. [54]

    Ronneberger, P

    O. Ronneberger, P. Fischer and T. Brox, U-net: Convolutional networks for biomedical image segmentation, in Medical image computing and computer-assisted intervention–MICCAI 2015 , pp. 234–241, 2015, DOI

    2015

  47. [55]

    Laureijs, J

    R. Laureijs, J. Amiaux, S. Arduini, J.-L. Augueres, J. Brinchmann, R. Cole et al., Euclid definition study report , ArXiv e-prints (2011) [ 1110.3193]

    2011 arXiv

  48. [56]

    Spergel, N

    D. Spergel, N. Gehrels, C. Baltay, D. Bennett, J. Breckinridge, M. Donahue et al., Wide-field infrarred survey telescope-astrophysics focused telescope assets WFIRST-AFTA 2015 report , ArXiv e-prints (2015) [ 1503.03757]. – 26 –

    2015 arXiv

  49. [57]

    Ivezi´ c, S.M

    ˇZ. Ivezi´ c, S.M. Kahn, J.A. Tyson, B. Abel, E. Acosta, R. Allsman et al., LSST: from science drivers to reference design and anticipated data products , ApJ 873 (2019) 111

    2019

  50. [58]

    Lintott, K

    C.J. Lintott, K. Schawinski, A. Slosar, K. Land, S. Bamford, D. Thomas et al., Galaxy Zoo: morphologies derived from visual inspection of galaxies from the Sloan Digital Sky Survey , MNRAS 389 (2008) 1179

    2008

  51. [59]

    Annis, M

    J. Annis, M. Soares-Santos, M.A. Strauss, A.C. Becker, S. Dodelson, X. Fan et al., The sloan digital sky survey coadd: 275 deg2 of deep sloan digital sky survey imaging on stripe 82 , ApJ 794 (2014) 120

    2014

  52. [60]

    Le Cun, L.D

    Y. Le Cun, L.D. Jackel, B. Boser, J.S. Denker, H.P. Graf, I. Guyon et al., Handwritten digit recognition: applications of neural network chips and automatic learning , IEEE Communications Magazine 27 (1989) 41

    1989

  53. [61]

    Goodfellow, Y

    I. Goodfellow, Y. Bengio and A. Courville, Deep learning, MIT press (2016)

    2016

  54. [62]

    LeCun, L

    Y. LeCun, L. Bottou, Y. Bengio and P. Haffner, Gradient-based learning applied to document recognition, Proc. IEEE 86 (1998) 2278

    1998

  55. [63]

    J. Gu, Z. Wang, J. Kuen, L. Ma, A. Shahroudy, B. Shuai et al., Recent advances in convolutional neural networks, Pattern Recognition 77 (2018) 354

    2018

  56. [64]

    K. He, X. Zhang, S. Ren and J. Sun, Deep residual learning for image recognition , in 2016 IEEE Conference on Computer Vision and Pattern Recognition , pp. 770–778, 2016, DOI

    2016

  57. [65]

    Tan and Q

    M. Tan and Q. Le, EfficientNet: Rethinking model scaling for convolutional neural networks , in Proceedings of the 36th International Conference on Machine Learning , vol. 97, pp. 6105–6114, 2019

    2019

  58. [66]

    Sandler, A

    M. Sandler, A. Howard, M. Zhu, A. Zhmoginov and L.-C. Chen, MobileNetV2: Inverted residuals and linear bottlenecks, in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4510–4520, 2018, DOI

    2018

  59. [67]

    J. Hu, L. Shen and G. Sun, Squeeze-and-excitation networks, in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 7132–7141, 2018, DOI

    2018

  60. [68]

    Tan and Q

    M. Tan and Q. Le, EfficientNetV2: Smaller models and faster training , in Proceedings of the 38th International Conference on Machine Learning , vol. 139, pp. 10096–10106, 2021

    2021

  61. [69]

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li and F.-F. Li, ImageNet: A large-scale hierarchical image database, in 2009 IEEE Conference on Computer Vision and Pattern Recognition , pp. 248–255, 2009, DOI

    2009

  62. [70]

    Walmsley, C

    M. Walmsley, C. Allen, B. Aussel, M. Bowles, K. Gregorowicz, I.V. Slijepcevic et al., Zoobot: Adaptable deep learning models for galaxy morphology , Journal of Open Source Software 8 (2023) 5312

    2023

  63. [71]

    Huang, Z

    G. Huang, Z. Liu, L. van der Maaten and K.Q. Weinberger, Densely connected convolutional networks, in 2017 IEEE Conference on Computer Vision and Pattern Recognition , pp. 4700–4708, 2017, DOI

    2017

  64. [72]

    Walmsley, C

    M. Walmsley, C. Lintott, T. G´ eron, S. Kruk, C. Krawczyk, K.W. Willett et al., Galaxy Zoo DECaLS: Detailed visual morphology measurements from volunteers and deep learning for 314 000 galaxies , MNRAS 509 (2022) 3966

    2022

  65. [73]

    Walmsley, A.M.M

    M. Walmsley, A.M.M. Scaife, C. Lintott, M. Lochner, V. Etsebeth, T. G´ eron et al., Practical galaxy morphology tools from deep supervised representation learning , MNRAS 513 (2022) 1581

    2022

  66. [74]

    Etsebeth, M

    V. Etsebeth, M. Lochner, M. Walmsley and M. Grespan, Astronomaly at scale: searching for anomalies amongst 4 million galaxies , MNRAS 529 (2024) 732. – 27 –

    2024

  67. [75]

    J.J. Popp, H. Dickinson, S. Serjeant, M. Walmsley, D. Adams, L. Fortson et al., Transfer learning for galaxy feature detection: Finding giant star-forming clumps in low-redshift galaxies using faster region-based convolutional neural network , RASTI 3 (2024) 174

    2024

  68. [76]

    Agarap, Deep learning using rectified linear units (ReLU) , ArXiv e-prints (2018) [1803.08375]

    A.F. Agarap, Deep learning using rectified linear units (ReLU) , ArXiv e-prints (2018) [1803.08375]

    2018 arXiv

  69. [77]

    J. Bridle, Training stochastic model recognition algorithms as networks can lead to maximum mutual information estimation of parameters , in Advances in Neural Information Processing Systems, vol. 2, 1989

    1989

  70. [78]

    Chollet et al., Keras, 2015

    F. Chollet et al., Keras, 2015

    2015

  71. [79]

    Abadi, A

    M. Abadi, A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro et al., Tensorflow: Large-scale machine learning on heterogeneous distributed systems , ArXiv e-prints (2016) [ 1603.04467]

    2016 arXiv

  72. [80]

    Vingelmann and F.H.P

    NVIDIA, P. Vingelmann and F.H.P. Fitzek, CUDA, release: 10.2.89 , 2020

    2020

  73. [81]

    Kingma and J

    D.P. Kingma and J. Ba, Adam: A method for stochastic optimization , ArXiv e-prints (2017) [1412.6980]

    2017 arXiv

  74. [82]

    Chattopadhay, A

    A. Chattopadhay, A. Sarkar, P. Howlader and V.N. Balasubramanian, Grad-CAM++: Generalized gradient-based visual explanations for deep convolutional networks , in 2018 IEEE Winter Conference on Applications of Computer Vision , pp. 839–847, 2018, DOI

    2018

  75. [83]

    Smilkov, N

    D. Smilkov, N. Thorat, B. Kim, F. Vi´ egas and M. Wattenberg,SmoothGrad: removing noise by adding noise , ArXiv e-prints (2017) [ 1706.03825]

    2017 arXiv

  76. [84]

    Gordon, A

    A.J. Gordon, A. Ferguson and R.G. Mann, Uncovering tidal treasures: Automated classification of faint tidal features in DECaLS data , MNRAS 534 (2024) 1459

    2024

  77. [85]

    Medina-Rosales, G

    E. Medina-Rosales, G. Cabrera-Vives and C.J. Miller, Mitigating bias in deep learning: training unbiased models on biased data for the morphological classification of galaxies , MNRAS 531 (2024) 52

    2024

  78. [86]

    Bhambra, B

    P. Bhambra, B. Joachimi and O. Lahav, Explaining deep learning of galaxy morphology with saliency mapping , MNRAS 511 (2022) 5032

    2022

  79. [87]

    Bamford, R.C

    S.P. Bamford, R.C. Nichol, I.K. Baldry, K. Land, C.J. Lintott, K. Schawinski et al., Galaxy Zoo: the dependence of morphology and colour on environment , MNRAS 393 (2009) 1324

    2009

  80. [88]

    Kelvin, S.P

    L.S. Kelvin, S.P. Driver, A.S.G. Robotham, E.N. Taylor, A.W. Graham, M. Alpaslan et al., Galaxy And Mass Assembly (GAMA): stellar mass functions by Hubble type , MNRAS 444 (2014) 1647

    2014

  81. [89]

    Mu˜ noz-Mateos, K

    J.C. Mu˜ noz-Mateos, K. Sheth, M. Regan, T. Kim, J. Laine, S. Erroz-Ferrer et al., The Spitzer Survey of stellar structure in galaxies (S4G): stellar masses, sizes, and radial profiles for 2352 nearby galaxies, ApJSS 219 (2015) 3

    2015

  82. [90]

    Kendall, C

    S. Kendall, C. Clarke and R.C.J. Kennicutt, Spiral structure in nearby galaxies – II. Comparative analysis and conclusions , MNRAS 446 (2015) 4155

    2015

  83. [91]

    Chang, A

    Y.-Y. Chang, A. van der Wel, E. da Cunha and H.-W. Rix, Stellar masses and star formation rates for 1 M galaxies from SDSS+WISE , ApJSS 219 (2015) 8

    2015

  84. [92]

    Porter-Temple, B.W

    R. Porter-Temple, B.W. Holwerda, A.M. Hopkins, L.E. Porter, C. Henry, T. Geron et al., Galaxy And Mass Assembly: Galaxy Zoo spiral arms and star formation rates , MNRAS 515 (2022) 3875

    2022

  85. [93]

    de Jong, G.A.V

    J.T.A. de Jong, G.A.V. Kleijn, T. Erben, H. Hildebrandt, K. Kuijken, G. Sikkema et al., The third data release of the Kilo-Degree Survey and associated data products , A&A 604 (2017) A134

    2017

  86. [94]

    Da Cunha, S

    E. Da Cunha, S. Charlot and D. Elbaz, A simple model to interpret the ultraviolet, optical and infrared emission from galaxies , MNRAS 388 (2008) 1595. – 28 –

    2008

  87. [95]

    Kauffmann, T.M

    G. Kauffmann, T.M. Heckman, S.D.M. White, S. Charlot, C. Tremonti, J. Brinchmann et al., Stellar masses and star formation histories for 105 galaxies from the Sloan Digital Sky Survey , MNRAS 341 (2003) 33

    2003

  88. [96]

    Brinchmann, S

    J. Brinchmann, S. Charlot, S.D.M. White, C. Tremonti, G. Kauffmann, T. Heckman et al., The physical properties of star-forming galaxies in the low-redshift universe , MNRAS 351 (2004) 1151

    2004

  89. [97]

    Tremonti, T.M

    C.A. Tremonti, T.M. Heckman, G. Kauffmann, J. Brinchmann, S. Charlot, S.D.M. White et al., The origin of the mass-metallicity relation: insights from 53,000 star-forming galaxies in the Sloan Digital Sky Survey , ApJ 613 (2004) 898

    2004

  90. [98]

    Kroupa, On the variation of the initial mass function , MNRAS 322 (2001) 231

    P. Kroupa, On the variation of the initial mass function , MNRAS 322 (2001) 231

    2001

  91. [99]

    Chakrabarti, G

    S. Chakrabarti, G. Laughlin and F.H. Shu, Branch, spur, and feather formation in spiral galaxies, ApJ 596 (2003) 220. – 29 –

    2003

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.