Gaussian Process Classification for Galaxy Blend Identification in LSST

Buchanan, J.; Schneider, Michael; Armstrong, R.; Muyskens, Amanda; Priest, Benjamin W.; Dana, Ryan

doi:10.3847/1538-4357/ac35ca

Cited by 7 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A compelling variant of GANs is also introduced and tested on nearby SDSS galaxies by Reiman & Gohre (2019), where they have the generator output two cutouts given one composite input cutout and fine-tune it by giving these two deblended cutouts as well as the two that were used to construct the composite to the discriminator. While in this work we improved the resolution/noise/pixel scale of ground-based images by training from the space data using GANs and performed source detection on the enhanced products using a standard peak-finding routine, Buchanan et al (2022) showed that detection of blends can be further improved using either Gaussian process or CNN-based blend classifiers. They also showed that the improvement in blend identification depends strongly on the footprint image they start with.…”

Section: Discussionmentioning

confidence: 99%

“…We simulated the blend sample from real galaxy cutouts, as opposed to using Galsim (Rowe et al 2015) models based on real galaxy distributions (e.g., Arcelin et al 2020;Buchanan et al 2022;Wang et al 2022) to be more representative of real galaxies. We restricted the training sample to galaxies brighter than the 24th magnitude in F750W, and in testing went down to 25th magnitude.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Deblending Galaxies with Generative Adversarial Networks

Hemmati

Huff

Nayyeri

et al. 2022

ApJ

View full text Add to dashboard Cite

Deep generative models including generative adversarial networks (GANs) are powerful unsupervised tools in learning the distributions of data sets. Building a simple GAN architecture in PyTorch and training on the CANDELS data set, we generate galaxy images with the Hubble Space Telescope (HST) resolution starting from a noise vector. We proceed by modifying the GAN architecture to improve Subaru Hyper Suprime-Cam (HSC) ground-based images by increasing their resolution to the HST resolution. We use the super-resolution GAN on a large sample of blended galaxies, which we create using CANDELS cutouts. In our simulated blend sample, ∼20% would unrecognizably be blended even in the HST-resolution cutouts. In the HSC-like cutouts this fraction rises to ∼90%. With our modified GAN we can lower this value to ∼50%. We quantify the blending fraction in the high, low, and GAN resolutions over the whole manifold of angular separation, flux ratios, sizes, and redshift difference between the two blended objects. The two peaks found by the GAN deblender result in improvement by a factor of 10 in the photometry measurement of the blended objects. Modifying the architecture of the GAN, we also train a multiwavelength GAN with HST cutouts in seven optical + near-infrared bands. This multiwavelength GAN improves the fraction of detected blends by another ∼10% compared to the single-band GAN. This is most beneficial to the current and future precision cosmology experiments (e.g., LSST, SPHEREx, Euclid, Roman), specifically those relying on weak gravitational lensing, where blending is a major source of systematic error.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Deblending Galaxies with Generative Adversarial Networks

Hemmati

Huff

Nayyeri

et al. 2022

ApJ

View full text Add to dashboard Cite

show abstract

“…A compelling variant of GANs is also introduced and tested on nearby SDSS galaxies by Reiman & Göhre 2019, where they have the generator output two cutouts given one composite input cutout and fine tune it by giving these two deblended cutouts as well as the two that were used to construct the composite to the discriminator. While in this work, we improved the resolution/noise/pixel scale of ground-based images by training from the space data using GANs and did source detection on the enhanced products using a standard peakfinding routine, Buchanan et al 2022 showed that detection of blends can be further improved using either Gaussian Process or CNN-based blend classifiers. They also showed that the improvement in blend identification depends strongly on the footprint image they start with.…”

Section: Discussionmentioning

confidence: 99%

“…We simulated the blend sample from real galaxy cutouts, as opposed to using Galsim (Rowe et al 2015) models based on real galaxy distributions (e.g., Arcelin et al 2020, Buchanan et al 2022, Wang et al 2022) to be more representative of real galaxies. We restricted the training sample to galaxies brighter than the 24th magnitude in F750W and in testing went down to 25th magnitude.…”

Section: Discussionmentioning

confidence: 99%

Deblending Galaxies with Generative Adversarial Networks

Hemmati,

Huff,

Nayyeri

et al. 2022

Preprint

View full text Add to dashboard Cite

Deep generative models including generative adversarial networks (GANs) are powerful unsupervised tools in learning the distributions of data sets. Building a simple GAN architecture in PyTorch and training on the CANDELS data set, we generate galaxy images with the Hubble Space Telescope resolution starting from a noise vector. We proceed by modifying the GAN architecture to improve the Subaru Hyper Suprime-Cam ground-based images by increasing their resolution to the HST resolution. We use the super resolution GAN on a large sample of blended galaxies which we create using CANDELS cutouts. In our simulated blend sample, ∼ 20% would unrecognizably be blended even in the HST resolution cutouts. In the HSC-like cutouts this fraction rises to ∼ 90%. With our modified GAN we can lower this value to ∼ 50%. We quantify the blending fraction in the high, low and GAN resolutions over the whole manifold of angular separation, flux ratios, sizes and redshift difference between the two blended objects. The two peaks found by the GAN deblender result in ten times improvement in the photometry measurement of the blended objects. Modifying the architecture of the GAN, we also train a Multi-wavelength GAN with seven band optical+NIR HST cutouts. This multi-wavelength GAN improves the fraction of detected blends by another ∼ 10% compared to the single-band GAN. This is most beneficial to the current and future precision cosmology experiments (e.g., LSST, SPHEREx, Euclid, Roman), specifically those relying on weak gravitational lensing, where blending is a major source of systematic error.

show abstract

“…A Gaussian process classifier was used for the classification, as research has shown its ability to predict well-calibrated probabilities and its superior performance compared to logistic regression [38]. Further, the Gaussian process classifier has been successfully used in medical studies [39,40].…”

Section: Feature Set and Modellingmentioning

confidence: 99%

Machine Learning and Explainable Artificial Intelligence using Counterfactual Explanations for Evaluating Posture Parameters

Dindorf¹,

Ludwig²,

Simon³

et al. 2023

Preprint

View full text Add to dashboard Cite

Postural deficits such as hyperlordosis (hollow back) or hyperkyphosis (hunchback) are relevant health issues. Diagnoses depend on the experience of the examiner and are therefore often subjective and prone to errors. Machine learning (ML) methods in combination with explainable ar-tificial intelligence (XAI) tools have proven useful for providing an objective, data-based orien-tation. However, only a few works have considered posture parameters, leaving the potential of more human-friendly XAI interpretations still untouched. Therefore, the present work proposes an objective, data-driven ML system for medical decision support that enables especially human-friendly interpretations using counterfactual explanations (CFs). Posture data for 1151 subjects were recorded by means of stereophotogrammetry. An expert-based classification of the subjects regarding the presence of hyperlordosis or hyperkyphosis was initially performed. Using a Gaussian progress classifier, the models were trained and interpreted using CFs. Label errors were flagged and re-evaluated using confident learning. Very good classification performances for both hyperlordosis and hyperkyphosis were found, whereby the re-evaluation and correction of the test labels led to a significant improvement (MPRAUC = 0.97). A statistical evaluation showed that the CFs seemed to be plausible in general. In the context of personalized medicine, the present study’s approach could be of importance for reducing diagnostic errors and thereby improving the individual adaptation of therapeutic measures. Likewise, it could be a basis for the development of apps for preventive posture assessment.

show abstract

Gaussian Process Classification for Galaxy Blend Identification in LSST

Cited by 7 publications

References 28 publications

Deblending Galaxies with Generative Adversarial Networks

Deblending Galaxies with Generative Adversarial Networks

Deblending Galaxies with Generative Adversarial Networks

Machine Learning and Explainable Artificial Intelligence using Counterfactual Explanations for Evaluating Posture Parameters

Contact Info

Product

Resources

About