The DOI matches its Crossref record; 61 of the 61 references it cites were found in a registry or an index; 7 of the 17 sources its claims cite were checked for support.
Self-supervised Representation Learning for Astronomical Images
Open accessAbstract
Abstract Sky surveys are the largest data generators in astronomy, making automated tools for extracting meaningful scientific information an absolute necessity. We show that, without the need for labels, self-supervised learning recovers representations of sky survey images that are semantically useful for a variety of scientific tasks. These representations can be directly used as features, or fine-tuned, to outperform supervised methods trained only on labeled data. We apply a contrastive learning framework on multiband galaxy photometry from the Sloan Digital Sky Survey (SDSS), to learn image representations. We then use them for galaxy morphology classification and fine-tune them for photometric redshift estimation, using labels from the Galaxy Zoo 2 data set and SDSS spectroscopy. In both downstream tasks, using the same learned representations, we outperform the supervised state-of-the-art results, and we show that our approach can achieve the accuracy of supervised models while using 2–4 times fewer labels for training. The codes, trained models, and data can be found at https://portal.nersc.gov/project/dasrepo/self-supervised-learning-sdss .
Overview
CoChat Summary
TL;DR
Self-supervised contrastive learning on unlabeled SDSS galaxy images learns representations that are useful for morphology classification and photometric redshift estimation. The learned representations outperform supervised baselines and can match supervised accuracy with 2–4× fewer labels.
Key findings
- Using the same learned representations, the authors outperform supervised state-of-the-art results on galaxy morphology classification and photometric redshift estimation.
- For photometric redshift estimation, pre-training on unlabeled data gives accuracy equivalent to supervised training on 2–4× more spectroscopic redshift labels.
- For the first Galaxy Zoo 2 morphology question, roughly a factor of 16 more labels are required in the supervised setting to achieve the same performance as the fine-tuned self-supervised model.
- In augmentation ablation, Gaussian noise was the strongest augmentation and PSF the weakest; best performance was achieved using all augmentations except PSF.
Methods
- Contrastive self-supervised learning with an InfoNCE loss on ∼1.2 million SDSS ugriz galaxy images of 64×64 pixels, using a ResNet50 encoder and augmentations including galactic extinction, PSF smoothing, rotation, jitter/crop, and Gaussian noise.
- Downstream tasks: Galaxy Zoo 2 morphology classification using 183,929 galaxies with both ugriz images and crowd-sourced classifications, and photometric redshift estimation using 517,190 labeled spectroscopic galaxies, with 400,000 for training and 103,000 for validation.
- Morphology classifiers included a supervised CNN from scratch, a linear classifier on frozen self-supervised representations, and a fine-tuned self-supervised encoder; photometric redshift used 180 redshift bins spanning 0 ≤ z ≤ 0.4 and metrics Δz, σ_MAD, and outlier percentage η.
Limitations
- The paper states that exact quantitative comparison to Domínguez Sánchez et al. (2018) and Walmsley et al. (2020) is not possible due to inconsistent datasets, and that augmentation findings may not generalize to other surveys with different resolutions, signal-to-noise ratios, or target objects.
- The paper states that spectroscopic labels impose selection bias: galaxies with specz labels are biased toward nearby bright objects with low galactic extinction and are not representative of all photometric galaxies.
- The text also makes evident that images span 25.3 arcseconds, which is not large enough for very nearby large galaxies, and that misclassifications arise from imaging artifacts and label uncertainty.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
7 Valid · 1 Not valid · 4 Not checked · 74 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“For example, due to the series of flux and quality cuts applied when selecting spectroscopic targets (Strauss et al. 2002), galaxies with specz labels have a distinct bias towards nearby bright objects with low galactic extinction.”
What the sources say
- Partly supportedcites Strauss et al. 2002 sourceExperimental — accuracy not yet measured
Section 4.1 (page unavailable) describes an r-band magnitude limit of 17.77, and §5.2 reports a median redshift of 0.104; the paper also describes photometric quality cuts. These support selection toward apparently bright, relatively nearby galaxies. But §4.1 says magnitudes are corrected for Galactic extinction before selection, and the cited passages do not establish a distinct bias toward low-extinction sightlines. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and we show that our approach can achieve the accuracy of supervised models while using 2-4 times fewer labels for training.”
- Also stated in: § 3.3. Photometric Redshift Estimation · p. 7 · boosted
“Impressively, the accuracy gained from pre-training on unlabeled data is equivalent to supervised training on 2 -4 × more specz labels, with no modifications in architecture size or complexity.”
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.3. Photometric Redshift Estimation · p. 7 · boosted
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Likewise, we found that classification became a straightforward task when using any subset of the labels.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Here we further show that the UMAP performed on the representations surpasses the utility of a UMAP performed directly on the images”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As evidenced by Figure 2, the self-supervised representations have achieved a high degree of separation between numerous types of galaxy morphologies.”
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
40 more claims
Create a free account to see all 86 claims and to verify the 1 reference still open.
- Probably not a substantial claim. 25
- It may state two separate claims as one. 18
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Another benefit to self-supervised models, other than training on unlabelled data, is the computational cost, self-supervised models have been shown to need only 1% of the computational power needed by supervised models (Hayat et al. 2021).”
From Detecting galaxy tidal features using self-supervised representation learning (2024) · cites this paper as Hayat et al. 2021
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Here, the authors proposed a two-step approach, first training in complete self-supervision (without needing any spectroscopic redshifts) a 1d encoding of galaxies postage stamps. And in a second step, training in a supervised way a shallow MLP to predict redshifts on a small training set of available spectroscopic redshifts.”
From The Dawes Review 10: The impact of deep learning for the analysis of galaxy surveys (2023) · cites this paper as Hayat et al. (2021)
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“They find for instance that their selfsupervised approach using 20% of labelled data achieves similar accuracy to a fully supervised training using labels for the entire dataset.”
From The Dawes Review 10: The impact of deep learning for the analysis of galaxy surveys (2023) · cites this paper as Hayat et al. (2021)
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The first work exploring self-supervised learning is by Hayat et al. (2021).”
From The Dawes Review 10: The impact of deep learning for the analysis of galaxy surveys (2023) · cites this paper as Hayat et al. (2021)
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Hayat et al. (2021) and Stein et al. (2022) show that once the encoder section of a self-supervised model has been trained one can conduct a simple similarity search to find objects of interest, or apply a linear classifier directly onto the encoder's outputs to separate the data into a variety of classes.”
From Detecting galaxy tidal features using self-supervised representation learning (2024) · cites this paper as Hayat et al. 2021
6 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ C.3. Performance metrics & additional results · p. 17
“Figure 11. Predicted labels compared to crowd-sourced answers for the first GZ2 question: 'Could this be a disk viewed edge-on?'.”
§ C.3. Performance metrics & additional results · p. 16
“Figures 11 and 12 show the classification results in the same style as Figure 3, but for the second and third GZ questions, respectively.”
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325
A simple framework for contrastive learning of visual representations
Chen, T., Kornblith, S., Norouzi, M., & Hinton, G. (2020) · arXiv · doi:10.48550/arxiv.2002.05709 · Cited by 10,857
MEASURING REDDENING WITH SLOAN DIGITAL SKY SURVEY STELLAR SPECTRA AND RECALIBRATING SFD
Schlafly, E. F., & Finkbeiner, D. P. (2011) · The Astrophysical Journal · doi:10.1088/0004-637x/737/2/103 · Cited by 7,531
UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
McInnes, L., Healy, J., & Melville, J. (2018) · arXiv · doi:10.48550/arxiv.1802.03426 · Cited by 7,101
Representation Learning with Contrastive Predictive Coding
Oord, A. V. D., Li, Y., & Vinyals, O. (2018) · arXiv · doi:10.48550/arxiv.1807.03748 · Cited by 6,242
Dimensionality Reduction by Learning an Invariant Mapping
Hadsell, R., Chopra, S., & LeCun, Y. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06) · doi:10.1109/cvpr.2006.100 · Cited by 5,336
LSST: From Science Drivers to Reference Design and Anticipated Data Products
Ivezić, Ž., Kahn, S. M., Tyson, J. A., Abel, B., Acosta, E., Allsman, R., Alonso, D., AlSayyad, Y., Anderson, S. F., Andrew, J., et al. (2019) · The Astrophysical Journal · doi:10.3847/1538-4357/ab042c · Cited by 4,357 · cited 3×
“In the near future, incoming sky surveys such as the Vera Rubin Observatory (Ivezić et al. 2019b), Euclid (Laureijs et al. 2011), the Nancy Grace Roman Space Telescope (Spergel et al. 2013), or the Square Kilometer Array6 will open yet another research epoch, where data sets are of sizes that…”
BackgroundMethodLSST: from Science Drivers to Reference Design and Anticipated Data Products
Ivezić, Ž., Kahn, S. M., Tyson, J. A., Abel, B., Acosta, E., Allsman, R., Alonso, D., AlSayyad, Y., Anderson, S. F., Andrew, J., et al. (2008) · arXiv · doi:10.48550/arxiv.0805.2366 · Cited by 4,357
THE ELEVENTH AND TWELFTH DATA RELEASES OF THE SLOAN DIGITAL SKY SURVEY: FINAL DATA FROM SDSS-III
Alam, S., Albareti, F. D., Prieto, C. A., Anders, F., Anderson, S. F., Anderton, T., Andrews, B. H., Armengaud, E., Aubourg, É., Bailey, S., et al. (2015) · The Astrophysical Journal Supplement Series · doi:10.1088/0067-0049/219/1/12 · Cited by 2,449 · cited 1×
“Our database of galaxies is assembled from Data Release 12 (DR12; Alam et al. 2015) of the SDSS.”
Method
Cited by
The Dawes Review 10: The impact of deep learning for the analysis of galaxy surveys
Huertas-Company, M., & Lanusse, F. (2023) · Publications of the Astronomical Society of Australia · doi:10.1017/pasa.2022.55 · Cited by 102
“The first work exploring self-supervised learning is by Hayat et al. (2021).”
BackgroundMethodInfluentialAstronomia ex machina: a history, primer and outlook on neural networks in astronomy
Smith, M. J., & Geach, J. E. (2023) · Royal Society Open Science · doi:10.1098/rsos.221454 · Cited by 111
“[231] trained SimCLR [185] on multi-band galaxy photometry from the SDSS [74].”
MethodAstroCLIP: a cross-modal foundation model for galaxies
Parker, L., Lanusse, F., Golkar, S., Sarra, L., Cranmer, M., Bietti, A., Eickenberg, M., Krawezik, G. P., McCabe, M., Morel, R., et al. (2024) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stae1450 · Cited by 192
Practical galaxy morphology tools from deep supervised representation learning
Walmsley, M., Scaife, A. M. M., Lintott, C. J., Lochner, M., Etsebeth, V., Géron, T., Dickinson, H. J., Fortson, L., Kruk, S., Masters, K. L., et al. (2022) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stac525 · Cited by 111
Mining for Strong Gravitational Lenses with Self-supervised Learning
Stein, G., Blaum, J., Harrington, P., Medan, T., & Lukić, Z. (2022) · The Astrophysical Journal · doi:10.3847/1538-4357/ac6d63 · Cited by 52
“…orientation of the galaxies, the number of sources, the clustering, and the color, while removing the noise and symmetries described by the data augmentations (for a visualization of how galaxy properties are organized in the representation space of a self-supervised model, see Hayat et al. 2021).”
BackgroundMethodInfluentialSelf-citationKodra, D., Andrews, B. H., Newman, J. A., Finkelstein, S. L., Fontana, A., Hathi, N. P., Salvato, M., Wiklind, T., Wuyts, S., Broussard, A., et al. (2023) · The Astrophysical Journal · doi:10.3847/1538-4357/ac9f12 · Cited by 49
“ML methods outperform template-based methods in the regime of complete and representative coverage of color-z space, such as Sloan Digital Sky Survey photo-z efforts (e.g., Beck et al. 2016; Pasquet et al. 2019; Hayat et al. 2021; Dey et al. 2022a), where datadriven interpolation is very accurate.”
MethodA review of unsupervised learning in astronomy
Fotopoulou, S. (2024) · Astronomy and Computing · doi:10.1016/j.ascom.2024.100851 · Cited by 68
Towards an astronomical foundation model for stars with a transformer-based model
Leung, H. W., & Bovy, J. (2023) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stad3015 · Cited by 80
Vega-Ferrero, J., Costantin, L., Pérez‐González, P. G., Sarmiento, R., Kartaltepe, J. S., Pillepich, A., Bagley, M. B., Finkelstein, S. L., McGrath, E. J., Knapen, J. H., et al. (2024) · The Astrophysical Journal · doi:10.3847/1538-4357/ad05bb · Cited by 55
“To that purpose, we apply a novel data-driven approach based on contrastive learning ( Hayat et al. 2021; Sarmiento et al. 2021 ) to a mass-complete sample of JWST galaxies at z 3 observed within the Cosmic Evolution Early Release Science ( CEERS; Finkelstein et al. 2017, 2022, 2023 ) survey.”
MethodSlijepcevic, I. V., Scaife, A. M. M., Walmsley, M., Bowles, M., Wong, O. I., Shabala, S. S., & Tang, H. (2022) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stac1135 · Cited by 39
Self-supervised anomaly detection for new physics
Dillon, B. M., Mastandrea, R., & Nachman, B. P. (2022) · Physical review. D/Physical review. D. · doi:10.1103/physrevd.106.056005 · Cited by 33
“[12], and with constituent level jet data in Ref.”
Deep Learning of Sea Surface Temperature Patterns to Identify Ocean Extremes
Prochaska, J. X., Cornillon, P., & Reiman, D. (2021) · Remote Sensing · doi:10.3390/rs13040744 · Cited by 36
Slijepcevic, I. V., Scaife, A. M. M., Walmsley, M., Bowles, M., Wong, O. I., Shabala, S. S., & White, S. V. (2023) · RAS Techniques and Instruments · doi:10.1093/rasti/rzad055 · Cited by 43
“More recently, Hayat et al. (2021) and Stein et al. (2021a) have inspired our work with their successful applications of contrastive learning methods to galaxy morphology classification and gravitational lens/non-lens classification, respectively; Walmsley et al. (2022) show performance…”
BackgroundUnsupervised Galaxy Morphological Visual Representation with Deep Contrastive Learning
Wei, S., Li, Y., Lu, W., Li, N., Liang, B., Dai, W., & Zhang, Z. (2022) · Publications of the Astronomical Society of the Pacific · doi:10.1088/1538-3873/aca04e · Cited by 26
“Hayat et al. (2021) have applied CL to perform galaxy morphological classification and photometric redshift estimation on multiband galaxy photometry from SDSS.”
MethodCarter, J. T., Feddema, J., Kothe, D. B., Neely, R., Pruet, J., Stevens, R. L., Balaprakash, P., Beckman, P., Foster, I. T., Iskra, K. A., et al. (2023) · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · doi:10.2172/1986455 · Cited by 41
CNN photometric redshifts in the SDSS at r ≤ 20
Treyer, M. A., Ouahmed, R. A., Pasquet, J., Arnouts, S., Bertin, E., & Fouchez, D. (2023) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stad3171 · Cited by 29
A brief review of contrastive learning applied to astrophysics
Sarmiento, R., & Knapen, J. H. (2023) · RAS Techniques and Instruments · doi:10.1093/rasti/rzad028 · Cited by 38
Pearce-Casey, R., Nagam, B. C., Wilde, J., Busillo, V., Ulivi, L., Andika, I. T., Manjón-García, A., Leuzzi, L., Matavulj, P., Serjeant, S., et al. (2025) · Astronomy & Astrophysics · doi:10.1051/0004-6361/202453152 · Cited by 18
Data compression and inference in cosmology with self-supervised machine learning
Akhmetzhanova, A., Mishra-Sharma, S., & Dvorkin, C. (2023) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stad3646 · Cited by 36
Enabling unsupervised discovery in astronomical images through self-supervised representations
Mohale, K., & Lochner, M. (2024) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stae926 · Cited by 31
AstroM 3 : A Self-supervised Multimodal Model for Astronomy
Rizhko, M., & Bloom, J. S. (2025) · The Astronomical Journal · doi:10.3847/1538-3881/adcbad · Cited by 18
The ROAD to discovery: Machine-learning-driven anomaly detection in radio astronomy spectrograms
Mesarcik, M., Boonstra, A., Iacobelli, M., Ranguelova, E., de Laat, C. T. A. M., & van Nieuwpoort, R. V. (2023) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202347182 · Cited by 18
Transfer Learning and Deep Metric Learning for Automated Galaxy Morphology Representation
Variawa, M. Z., van Zyl, T. L., & Woolway, M. (2022) · IEEE Access · doi:10.1109/access.2022.3150881 · Cited by 16
“Metric learning also has applications in image recognition and multi-label classification [21]–[23] and has the potential to be applied to other astronomical use cases [24]–[26].”
BackgroundMethodXu, Q., Shen, S., de Souza, R. S., Chen, M., Ye, R., She, Y., Chen, Z., Ishida, É. E. O., Krone-Martins, A., & Durgesh, R. (2023) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stad3181 · Cited by 22
Detecting galaxy tidal features using self-supervised representation learning
Desmons, A., Brough, S., & Lanusse, F. (2024) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stae1402 · Cited by 15
Self-supervised similarity search for large scientific datasets
Stein, G., Harrington, P., Blaum, J., Medan, T., & Lukic, Z. (2021) · arXiv · doi:10.48550/arxiv.2110.13151 · Cited by 13
Self-citationTiming the last major merger of galaxy clusters with large halo sparsity
Richardson, T. R. G., & Corasaniti, P.-S. (2022) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stac1241 · Cited by 13
PICZL: Image-based photometric redshifts for AGN
Roster, W., Salvato, M., Krippendorf, S., Saxena, A., Shirley, R. A., Büchner, J., Wolf, J., Dwelly, T., Bauer, F. E., Aird, J., et al. (2024) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202452361 · Cited by 26
Xie, B., Yao, X., Mao, W., Rafiei, M. H., & Hu, N. (2022) · Computational Materials Science · doi:10.1016/j.commatsci.2022.111834 · Cited by 11
Bridging gaps with computer vision: AI in (bio)medical imaging and astronomy
Rezaei, S., Chegeni, A., Javadpour, A., Sadr, A. V., Cao, L., Röttgering, H. J. A., & Staring, M. (2024) · Astronomy and Computing · doi:10.1016/j.ascom.2024.100921 · Cited by 13
Dillmann, S., Martínez-Galarza, J. R., Soria, R., Stefano, R. D., & Kashyap, V. L. (2024) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stae2808 · Cited by 21
Aguilar-Argüello, G., Fuentes-Pineda, G., Hernández-Toledo, H. M., Martínez, L. A., Vázquez-Mata, J. A., Brough, S., Demarco, R., Ghosh, A., Jiménez-Teja, Y., Martin, G., et al. (2025) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/staf085 · Cited by 11
Ukraine, M. A. O. O. T. N. O., Vavilova, I. B., Khramtsov, V., Dobrycheva, D. V., Ukraine, M. A. O. O. T. N. O., Vasylenko, M. Y., Ukraine, M. A. O. O. T. N. O., Elyiv, A., Ukraine, M. A. O. O. T. N. O., Мelnyk, О. V., et al. (2022) · Kosmìčna nauka ì tehnologìâ · doi:10.15407/knit2022.01.003 · Cited by 11
“As for the combination of photometry and image galaxy SDSS data with Galaxy Zoo labels, we note the work by (59), who applied self-supervised representation learning.”
MethodMultimodality for improved CNN photometric redshifts
Ouahmed, R. A., Arnouts, S., Pasquet, J., Treyer, M. A., & Bertin, E. (2023) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202347395 · Cited by 14
Gondhalekar, Y., Chies-Santos, A. L., de Souza, R. S., Queiroz, C., Lopes, A. R., Ferrari, F., Azevedo, G. M., Monteiro-Pereira, H., Overzier, R., Castelli, A. V. S., et al. (2024) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stae1410 · Cited by 8
Jia, H., Zhu, H.-M., & Pen, U. (2023) · The Astrophysical Journal · doi:10.3847/1538-4357/aca8aa · Cited by 7
Jiménez, M. A., Alfaro, E. J., Torres, M. T., & Triguero, I. (2023) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stad2852 · Cited by 8
Radio Galaxy Zoo: Leveraging latent space representations from variational autoencoder
Andrianomena, S., & Tang, H. (2024) · Journal of Cosmology and Astroparticle Physics · doi:10.1088/1475-7516/2024/06/034 · Cited by 8
“[12] considered SimCLR method [13] to learn representations of astronomical images from SDSS, and [14] opted for Boostrap Your Own Latent ( BYOL ) method [8] to extract important features of radio galaxies.”
MethodFu, M., Song, Y., Lv, J., Cao, L., 贾, P. 鹏. J., Li, N., Li, X., Liu, J., Luo, A.-L., Qiu, B., et al. (2024) · Chinese Physics C · doi:10.1088/1674-1137/ad50ab · Cited by 8
Self-supervised learning on MeerKAT wide-field continuum images
Lastufka, E., Bait, O., Taran, O., Drozdova, M., Kinakh, V., Piras, D., Audard, M., Dessauges-Zavadsky, M., Holotyak, T., Schaerer, D., et al. (2024) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202449964 · Cited by 11
Neural network emulator for atmospheric chemical ODE
Liu, Z.-S., Clusius, P. S., & Boy, M. H. (2025) · Neural Networks · doi:10.1016/j.neunet.2024.107106 · Cited by 7
A survey on design choices for self-supervised learning in computer vision
Wittscher, L. (2026) · Artificial Intelligence Review · doi:10.1007/s10462-026-11506-9 · Cited by 5
Measuring the Intracluster Light Fraction with Machine Learning
Canepa, L., Brough, S., Lanusse, F., Montes, M., & Hatch, N. A. (2025) · The Astrophysical Journal · doi:10.3847/1538-4357/adabc7 · Cited by 10
CLAP: I. Resolving miscalibration for deep learning-based galaxy photometric redshift estimation
Lin, Q., Ruan, H., Fouchez, D., Chen, S., Li, R., Montero-Camacho, P., Napolitano, N. R., Ting, Y.-S., & Zhang, W. (2024) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202349113 · Cited by 5
Kerner, H., Rebbapragada, U. D., Wagstaff, K. L., et al. (2022) · Frontiers in Astronomy and Space Sciences · doi:10.3389/fspas.2022.867947 · Cited by 3
Gharat, S., Bose, B., Borthakur, A., et al. (2023) · RAS Techniques and Instruments · doi:10.1093/rasti/rzad027 · Cited by 2
Exploring galaxy properties of eCALIFA with contrastive learning
Martínez-Solaeche, G., García-Benito, R., Delgado, R. M. G., Díaz-García, L. A., Sánchez, S. F., Conrado, A., & Rodríguez-Martín, J. E. (2024) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202450074 · Cited by 6
Howie, S., Cheng, T.-Y., & Baugh, C. M. (2025) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/staf1926 · Cited by 3
A Multimodal Transfer Learning Method for Classifying Images of Celestial Point Sources
Wang, B., Hong, S., Yuan, Z., et al. (2023) · Publications of the Astronomical Society of the Pacific · doi:10.1088/1538-3873/acfbb9 · Cited by 2
“In addition, self-supervised learning is also a special kind of unsupervised learning ( Hayat et al. 2021 ) .”
BackgroundAstroMAE: Redshift Prediction Using a Masked Autoencoder with a Novel Fine-Tuning Architecture
Fathkouhi, A. D., & Fox, G. (2024) · IEEE International Conference on eScience · doi:10.1109/e-science62913.2024.10678724 · Cited by 5
“Hayat et al. [18] used contrastive learning for pretraining the ResNet50 model, with the pretrained weights later employed for morphology classification and redshift prediction.”
MethodThe Fundamental Patterns of Sea Surface Temperature
Prochaska, J. X., Guo, E., Cornillon, P., & Buckingham, C. E. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3300272 · Cited by 4
“within a massive imagery dataset (see [23]).”
Capturing the physics of MaNGA galaxies with self-supervised Machine Learning
Sarmiento, R., Huertas-Company, M., Knapen, J. H., Sánchez, S. F., Domínguez Sánchez, H., Drory, N., & Falcón-Barroso, J. (2021) · The Astrophysical Journal · doi:10.3847/1538-4357/ac1dac · Cited by 1
“Abul Hayat et al. (2021) and Chen et al. (2020b) found that this framework can obtain the same levels of accuracy for classification as a supervised approach but reduce the number of training labels.”
BackgroundThe Hubble Image Similarity Project
White, R. L., & Peek, J. E. G. (2025) · The Astronomical Journal · doi:10.3847/1538-3881/adcb43 · Cited by 1
“Hayat et al. (2021) applied self-supervised learning to images of galaxies to extract morphological information.”
Yin, X., Fang, G., Lu, S., Lin, Z., Dai, Y., & Zhou, C.-C. (2025) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202555584 · Cited by 1
Detecting model misspecification in cosmology with scale-dependent normalizing flows
Akhmetzhanova, A., Cuesta-Lazaro, C., & Mishra-Sharma, S. (2025) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ae107d · Cited by 1
Deep Learning Improves Photometric Redshifts in All Regions of Color Space
Moran, E. R., Andrews, B. H., Newman, J. A., & Dey, B. (2026) · The Astrophysical Journal · doi:10.3847/1538-4357/ae2ace · Cited by 5
“…mechanisms, a variety of modern deep learning models—including CNNs ( J. Pasquet et al. 2019 ) , semisupervised representation learning ( M. A. Hayat et al. 2021 ) , and deep capsule networks ( B. Dey et al. 2022a ) —have all significantly outperformed classical ML photo-z methods ( namely…”
Enhancing Ly α Emitter Identification in HETDEX with a Convolutional Neural Network
Mukae, S., Cooper, E. M., Gebhardt, K., Davis, D., House, L. R., Qezlou, M., Muñoz, J. B., Saito, S., Farrow, D. J., Gronwall, C. A., et al. (2026) · The Astronomical Journal · doi:10.3847/1538-3881/ae5baf · Cited by 3
Space Imaging Point Source Detection and Characterization
Ribeiro, F. S. F., Garcia, P. J. V., Silva, M., & Cardoso, J. S. (2024) · IEEE Access · doi:10.1109/access.2024.3420217 · Cited by 3
“This approach has already been implemented in space images [119], [120] but not for object detection.”
BackgroundMethodGentile, F., Tortora, C., Covone, G., Koopmans, L. V. E., Spiniello, C., Fan, Z., Li, R., Liu, D., Napolitano, N. R., Vaccari, M., et al. (2021) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stab3386 · Cited by 35
A self-supervised learning approach for astronomical images
Martinazzo, A. (2021) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/d.45.2021.tde-11012022-203357 · Cited by 1
Embedding physics into deep learning for modeling spatiotemporal systems
Ren, P. (2022) · Digital Repository Service (Northeastern University) · doi:10.17760/d20467268 · Cited by 1
Wu, J. F., Peek, J. E. G., Tollerud, E. J., Mao, Y.-Y., Nadler, E. O., Geha, M., Wechsler, R. H., Kallivayalil, N., & Weiner, B. J. (2022) · The Astrophysical Journal · doi:10.3847/1538-4357/ac4eea · Cited by 29
“Other approaches are better suited for regressing the redshifts of more-distant galaxies using photometric information and images (e.g., Pasquet et al. 2019; Hayat et al. 2021; Schuldt et al. 2021).”
Xie, B., Yao, X., Mao, W., Rafiei, M. H., & Hu, N. (2022) · Research Square · doi:10.21203/rs.3.rs-1241474/v1
CNN photometric redshifts in the SDSS at $r\leq 20$Preprint
Treyer, M. A., Ouahmed, R. A., Pasquet, J., Arnouts, S., Bertin, E., & Fouchez, D. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2310.02173
Heneka, C. S., Niebling, J., Tang, H., Balakrishnan, V., Gawlikowski, J., Kasieczka, G., Segal, G., Gan, H., & Chamarthi, S. (2024) · Astrophysics and space science library · doi:10.1007/978-3-031-58468-8_10
Enhanced Classification of Astronomical Images Using Optimized Neural Networks via NAS
Gharge, S. D., Katurde, A. D., Shinde, B. B., et al. (2025) · International Conference on Knowledge Engineering and Communication Systems (ICKECS) · doi:10.1109/ickecs65700.2025.11035681
Engel, A. W., Byler, N., Tsou, A., Narayan, G., Bonilla, E., & Smith, I. (2025) · The Astrophysical Journal · doi:10.3847/1538-4357/adf5b8 · Cited by 1
“…Neural Networks or CNNs) to estimate the redshift directly from cutouts, rather than from derived features available in catalogs (J. Pasquet et al. 2019; B. Dey et al. 2022a; M. A. Hayat et al. 2021; A. D’Isanto & K. L. Polsterer 2018; B. Hoyle 2016; B. Henghes et al. 2022; L. Parker et al. 2023).”
Buatthaisong, N., Slijepcevic, I. V., Scaife, A. M. M., Bowles, M., Hopkins, A., Mohan, D., Shabala, S. S., & Wong, O. I. (2025) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/staf1942 · Cited by 4
Scalable cosmic AI inference using cloud serverless computing
Staylor, M., Fathkouhi, A. D., Islam, M. K., O’Hara, K., Goudjil, R. G., Fox, G., & Fox, G. (2025) · The International Journal of High Performance Computing Applications · doi:10.1177/10943420251399942 · Cited by 1
“…models trained on large-scale astronomical image datasets have proven to be powerful tools for improving the accuracy and efficiency of tasks such as redshift prediction, morphology classification, and similarity search Hayat et al. (2021); Lanusse et al. (2023); Fathkouhi and Fox (2024).”
FAMA—A Scalable Foundational Astronomical Masked Autoencoder for Astronomical Image Analysis
Lv, J., Li, X., Cao, L., et al. (2026) · The Astrophysical Journal Supplement Series · doi:10.3847/1538-4365/ae4509
“We further compared FAMA against leading self-supervised frameworks, including contrastive learning approaches ( Stein, M. A. Hayat et al. 2021; AstroCLIP, F. Lanusse et al. 2023 ) and autoencoder-based methods ( LVM, M. Fu et al. 2024 ) .”
InfluentialGalaxy morphology with Euclid : simulations, forecasts and detection with deep learning
Bretonnière, H. (2022) · theses.fr (ABES) · doi:10.70675/8f129d60z09daz4317zb532z6ea391cbdd9f
Galaxy redshift estimation from multi-band images with deep learning : strengths and challenges
Ouahmed, M. R. A. (2023) · theses.fr (ABES) · doi:10.70675/687e0665z3cb0z4fadza33bz5a64b0fe417c
A Vision Transformer Adaptation of SimCLRv2 for Self-Supervised Galaxy Detection
Biswas, U., Azmi, A. A., Sarker, M. M., Afrin, T., Ahmed, F., & Islam, M. S. (2025) · IEEE International Conference on Signal Processing, Information, Communication and Systems (SPICSCON) · doi:10.1109/spicscon69221.2025.11504180
“More recent studies have shown that self-supervised learning can be utilized effectively for different tasks in astronomy, such as galaxy morphology classification and star-galaxy separation [13].”
Stellar Density Classification and Regression for CSST Multi-color Imaging Using Deep Learning
Lai, J., Lam, M. I., Chen, J., et al. (2026) · Publications of the Astronomical Society of the Pacific · doi:10.1088/1538-3873/ae6d75
“Recent studies have demonstrated ResNet-50’s effectiveness in astronomical regression tasks, including photometric redshift estimation (M. A. Hayat et al. 2021; V. Sandeep et al. 2021) and metallicity prediction (L. Monti et al. 2024).”
Self-Supervised Spectral Representation Learning for LAMOST
Zhang, W., Zhou, A., Yuan, L., et al. (2026) · Universe · doi:10.3390/universe12060181
“In astronomy, self-supervised learning has been widely applied to galaxy morphology classification [19,25], yet its application to spectral data remains limited.”
Contrastive learning of extragalactic stellar streams
Benítez, E. M., Mes, J., Miró-Carretero, J., Kuijken, K. H., & Helmi, A. (2026) · Astronomy and Astrophysics · doi:10.1051/0004-6361/202659211
“Figure 6 presents a two-dimensional UMAP projection of the NNCLR embedding vectors of the model trained with tiered-sigmoid-scaled images, arranged into a mosaic of image cutouts as per Hayat et al. (2021).”
Influential
Metadata
- Authors
- Hayat, M. A., Stein, G., Harrington, P., Lukić, Z., & Mustafa, M.
- Affiliations
- Lawrence Berkeley National Laboratory (US)
- Venue
- The Astrophysical Journal Letters
- Publisher
- American Astronomical Society
- Published
- 2021-04-01
- Volume
- 911
- Issue
- 2
- Pages
- L33
- ISSN
- 2041-8205 (ISSN-L)
- 2041-8213
- OpenAlex
- W3118065583
- Cited by
- 79
- References
- 50
- Citations by year
- 2026: 8 · 2025: 14 · 2024: 18 · 2023: 20 · 2022: 15 · 2021: 4
- FWCI
- 7.60
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0/ (vor, 25 d)
- iopscience.iop.org/info/page/text-and-data-mining (tdm, 25 d)
- Locations
- The Astrophysical Journal Letters (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion, cc-by)
- eScholarship (California Digital Library) (submittedVersion)
- arXiv (Cornell University)
- Funders
- NERSC (AC02-05CH11231)
- Topics
- Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Galaxies: Formation, Evolution, Phenomena
- Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Astronomical Observations and Instrumentation
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- self-supervised representation learning, contrastive learning, celestial imagery, sky survey, galaxy morphology classification, photometric redshift estimation, Galaxy Zoo 2, Sloan Digital Sky Survey, label efficiency
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.