Simulation-Based Inference for Probabilistic Galaxy Detection and Deblending
University of Maryland, College Park（马里兰大学帕克分校）；University of Michigan（密歇根大学）；Leinweber Institute for Theoretical Physics, University of Michigan（密歇根大学莱因韦伯理论物理研究所）；The Voleon Group（Voleon集团）；Carnegie Mellon University（卡内基梅隆大学）
Ismael Mendoza, Derek Hansen, Runjing Liu, Zhe Zhao, Ziteng Pang, Axel Guinot, Camille Avestruz, Jeffrey Regier, the LSST Dark Energy Science Collaboration
备注：
23 pages, 11 figures, 2 tables, Published in the Open Journal of Astrophysics
第四阶段暗能量宽视场调查，如 Vera C. Rubin Observatory 的 Legacy Survey of Space and Time（LSST），将观测到前所未有的星系密度。因此，大多数成像星系将视觉上重叠，这种现象称为混叠。混叠预计将成为天文学测量中的主要系统误差来源。为缓解这种系统误差，我们提出了一种新的概率方法用于检测、去混叠和测量星系的性质，称为贝叶斯光源分离器（BLISS）。给定一个天文调查图像，BLISS 使用卷积神经网络生成一个概率天文目录，通过近似光子数、光心位置和类型（星系 vs 星星）的后验分布。BLISS 还包括一个去噪自动编码器来重建未混叠的星系轮廓。作为演示 BLISS 在宇宙学应用可行性的一个初步步骤，我们将我们的方法应用于模拟的单波段图像，其特性代表了 LSST 第十年的 coadds。首先，我们研究每个 BLISS 组件独立地，并检查其概率输出作为信噪比和混叠程度函数。然后，通过将 BLISS 的概率检测传播到其去混叠器，我们生成每个物体的光度后验。使用这些后验会产生比确定性检测更大的改进，特别是在高度混叠和暗淡的物体上。这些结果突显了 BLISS 作为可扩展、具有不确定性的工具，用于缓解下一代宇宙学调查中混叠引起的系统误差的潜力。
英文摘要：
Stage-IV dark energy wide-field surveys, such as the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), will observe an unprecedented number density of galaxies. As a result, the majority of imaged galaxies will visually overlap, a phenomenon known as blending. Blending is expected to be a leading source of systematic error in astronomical measurements. To mitigate this systematic, we propose a new probabilistic method for detecting, deblending, and measuring the properties of galaxies, called the Bayesian Light Source Separator (BLISS). Given an astronomical survey image, BLISS uses convolutional neural networks to produce a probabilistic astronomical catalog by approximating the posterior distribution over the number of light sources, their centroids' locations, and their types (galaxy vs. star). BLISS additionally includes a denoising autoencoder to reconstruct unblended galaxy profiles. As a first step towards demonstrating the feasibility of BLISS for cosmological applications, we apply our method to simulated single-band images whose properties are representative of year-10 LSST coadds. First, we study each BLISS component independently and examine its probabilistic output as a function of SNR and degree of blending. Then, by propagating the probabilistic detections from BLISS to its deblender, we produce per-object flux posteriors. Using these posteriors yields a substantial improvement in aperture flux residuals relative to deterministic detections alone, particularly for highly blended and faint objects. These results highlight the potential of BLISS as a scalable, uncertainty-aware tool for mitigating blending-induced systematics in next-generation cosmological surveys.
We present a new neural network approach for deblending galaxy images in astronomical data using Residual Dense Neural network (RDN) architecture. We train the network on synthetic galaxy images similar to the typical arrangements of field galaxies with a finite point spread function (PSF) and realistic noise levels. The main novelty of our approach is the usage of two distinct neural networks: i) a deblending network which isolates a single galaxy postage stamp from the composite and, ii) a classifier network which counts the remaining number of galaxies. The deblending proceeds by iterativel…
Probabilistic cataloging (PCAT) outperforms traditional cataloging methods on single-band optical data in crowded fields (Portillo et al. 2017). We extend our work to multiple bands, achieving greater sensitivity ($\sim$ 0.4 mag) and greater speed (500x) compared to previous single-band results. We demonstrate the effectiveness of multiband PCAT on mock data, both in terms of recovering accurate posteriors in the catalog space, and in directly deblending sources. When applied to Sloan Digital Sky Survey (SDSS) observations of M2, taking Hubble Space Telescope data as truth, our joint fit on $r…
Near-future large galaxy surveys will encounter blended galaxy images at a fraction of up to 50% in the densest regions of the universe. Current deblending techniques may segment the foreground galaxy while leaving missing pixel intensities in the background galaxy flux. The problem is compounded by the diffuse nature of galaxies in their outer regions, making segmentation significantly more difficult than in traditional object segmentation applications. We propose a novel branched generative adversarial network (GAN) to deblend overlapping galaxies, where the two branches produce images of th…
We introduce a new algorithm for colour separation and deblending of multi-band astronomical images called MuSCADeT which is based on Morpho-spectral Component Analysis of multi-band images. The MuSCADeT algorithm takes advantage of the sparsity of astronomical objects in morphological dictionaries such as wavelets and their differences in spectral energy distribution (SED) across multi-band observations. This allows us to devise a model independent and automated approach to separate objects with different colours. We show with simulations that we are able to separate highly blended objects an…
Blending of galaxies has a major contribution in the systematic error budget of weak lensing studies, affecting photometric and shape measurements, particularly for ground-based, deep, photometric galaxy surveys, such as the Rubin Observatory Legacy Survey of Space and Time (LSST). Existing deblenders mostly rely on analytic modelling of galaxy profiles and suffer from the lack of flexible yet accurate models. We propose to use generative models based on deep neural networks, namely variational autoencoders (VAE), to learn probabilistic models directly from data. We train a VAE on images of ce…
We present a new probabilistic method for detecting, deblending, and cataloging astronomical sources called the Bayesian Light Source Separator (BLISS). BLISS is based on deep generative models, which embed neural networks within a Bayesian model. For posterior inference, BLISS uses a new form of variational inference known as Forward Amortized Variational Inference. The BLISS inference routine is fast, requiring a single forward pass of the encoder networks on a GPU once the encoder networks are trained. BLISS can perform fully Bayesian inference on megapixel images in seconds, and produces h…
We present an extension of the multi-band galaxy fitting method scarlet which allows the joint modeling of astronomical images from different instruments, by performing simultaneous resampling and convolution. We introduce a fast and formally accurate linear projection operation that maps a pixelated model at a given resolution onto an observation frame with a different point spread function and pixel scale. We test our implementation against the well-tested resampling and convolution method in galsim on simulated images mimicking observations with the Euclid space telescope and the Vera C. Ru…
Deep optical images are often crowded with overlapping objects. This is especially true in the cores of galaxy clusters, where images of dozens of galaxies may lie atop one another. Accurate measurements of cluster properties require deblending algorithms designed to automatically extract a list of individual objects and decide what fraction of the light in each pixel comes from each object. We present new software called the Gradient And INterpolation based deblender (GAIN) as a secondary deblender to improve deblending the images of cluster cores. This software relies on using image intensit…
ULTIMATE-deblending I. A 50-band Ultraviolet to Mid-infrared Photometric Catalog Combining Space- and Ground-based Data in the JWST/PRIMER survey
School of Astronomy and Space Science, Nanjing University（南京大学天文与空间科学学院）；Key Laboratory of Modern Astronomy and Astrophysics, Nanjing University, Ministry of Education（南京大学现代天文与天体物理教育部重点实验室）；Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM（巴黎萨克雷大学、巴黎西岱大学、法国原子能和替代能源委员会、法国国家科学研究中心、天体物理学信息与方法研究所）；INAF – Osservatorio Astronomico di Roma（意大利国家天体物理研究所罗马天文台）；Chinese Academy of Sciences South America Center for Astronomy, National Astronomical Observatories, CAS（中国科学院南美天文中心、中国科学院国家天文台）
We apply a new deep learning technique to detect, classify, and deblend sources in multi-band astronomical images. We train and evaluate the performance of an artificial neural network built on the Mask R-CNN image processing framework, a general code for efficient object detection, classification, and instance segmentation. After evaluating the performance of our network against simulated ground truth images for star and galaxy classes, we find a precision of 92% at 80% recall for stars and a precision of 98% at 80% recall for galaxies in a typical field with $\sim30$ galaxies/arcmin$^2$. We …
DeepDISC on LSST Data Preview 1: Scene-Level Detection, Deblending, and Star/Galaxy Classification in the ECDFS and EDFS Fields
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；National Center for Supercomputing Applications, University of Illinois Urbana-Champaign（美国国家超级计算应用中心）；Center for Artificial Intelligence Innovation, University of Illinois Urbana-Champaign（人工智能创新中心）
Encoder-Decoder networks such as U-Nets have been applied successfully in a wide range of computer vision tasks, especially for image segmentation of different flavours across different fields. Nevertheless, most applications lack of a satisfying quantification of the uncertainty of the prediction. Yet, a well calibrated segmentation uncertainty can be a key element for scientific applications such as precision cosmology. In this on-going work, we explore the use of the probabilistic version of the U-Net, recently proposed by Kohl et al (2018), and adapt it to automate the segmentation of gala…
Cataloging is challenging in crowded fields because sources are extremely covariant with their neighbors and blending makes even the number of sources ambiguous. We present the first optical probabilistic catalog, cataloging a crowded (~0.1 sources per pixel brighter than 22nd magnitude in F606W) Sloan Digital Sky Survey r band image from M2. Probabilistic cataloging returns an ensemble of catalogs inferred from the image and thus can capture source-source covariance and deblending ambiguities. By comparing to a traditional catalog of the same image and a Hubble Space Telescope catalog of the …
Argonne National Laboratory（阿贡国家实验室）；National Astronomical Observatories of China（中国科学院国家天文台）；University of California, Berkeley（加州大学伯克利分校）
Upcoming large astronomical surveys are expected to capture an unprecedented number of strong gravitational lensing systems. Deep learning is emerging as a promising practical tool for the detection and quantification of these galaxy-scale image distortions. The absence of large quantities of representative data from current astronomical surveys motivates the development of a robust forward-modeling approach using synthetic lensing images. Using a mock sample of strong lenses created upon a state-of-the-art extragalactic catalogs, we train a modular deep learning pipeline for uncertainty-quant…
We present a novel Probabilistic Flux Variation Gradient (PFVG) approach to to separate the contributions of active galactic nuclei (AGN) and host galaxies in the context of photometric reverberation mapping (PRM) of AGN. We explored the ability of recovering the fractional contribution in a model-independent way using the entire set of light curves obtained through different filters and photometric apertures simultaneously. The method is based on the observed bluer when brighter phenomenon that is attributed to the superimposition of a two-component structure; the red host galaxy, which is co…
Cross-matching catalogues from radio surveys to catalogues of sources at other wavelengths is extremely hard, because radio sources are often extended, often consist of several spatially separated components, and often no radio component is coincident with the optical/infrared host galaxy. Traditionally, the cross-matching is done by eye, but this does not scale to the millions of radio sources expected from the next generation of radio surveys. We present an innovative automated procedure, using Bayesian hypothesis testing, that models trial radio-source morphologies with putative positions o…
We invoke a Gaussian mixture model (GMM) to jointly analyse two traditional emission-line classification schemes of galaxy ionization sources: the Baldwin-Phillips-Terlevich (BPT) and $\rm W_{Hα}$ vs. [NII]/H$α$ (WHAN) diagrams, using spectroscopic data from the Sloan Digital Sky Survey Data Release 7 and SEAGal/STARLIGHT datasets. We apply a GMM to empirically define classes of galaxies in a three-dimensional space spanned by the $\log$ [OIII]/H$β$, $\log$ [NII]/H$α$, and $\log$ EW(H$α$), optical parameters. The best-fit GMM based on several statistical criteria suggests a solution around fou…
We demonstrate how galaxy morphologies can be represented by weighted sums of "eigengalaxies" and how eigengalaxies can be used in a probabilistic framework to enable principled and simplified approaches in a variety of applications. Eigengalaxies can be derived from a Principal Component Analysis (PCA) of sets of single- or multi-band images. They encode the image space equivalent of basis vectors that can be combined to describe the structural properties of large samples of galaxies in a massively reduced manner. As an illustration, we show how a sample of 10,243 galaxies in the Hubble Space…
DB-Bench: Benchmarking Deblenders for LSST DESC Using the Blending ToolKit
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；National Center for Supercomputing Applications, University of Illinois at Urbana-Champaign（美国国家超级计算应用中心，伊利诺伊大学厄巴纳-香槟分校）；Center for Artificial Intelligence Innovation, University of Illinois at Urbana-Champaign（人工智能创新中心，伊利诺伊大学厄巴纳-香槟分校）；Université Paris Cité（巴黎西岱大学）；CNRS（法国国家科学研究中心）；CEA（法国原子能和替代能源委员会）
Probabilistic redshift estimation of unresolved galaxies from multi-band background light maps
Ruhr University Bochum（鲁尔大学波鸿）；Stanford University（斯坦福大学）；SLAC National Accelerator Laboratory（斯坦福直线加速器中心）；The University of British Columbia（不列颠哥伦比亚大学）
Many scientific investigations of photometric galaxy surveys require redshift estimates, whose uncertainty properties are best encapsulated by photometric redshift (photo-z) posterior probability density functions (PDFs). A plethora of photo-z PDF estimation methodologies abound, producing discrepant results with no consensus on a preferred approach. We present the results of a comprehensive experiment comparing twelve photo-z algorithms applied to mock data produced for The Rubin Observatory Legacy Survey of Space and Time (LSST) Dark Energy Science Collaboration (DESC). By supplying perfect …
Analysis of galaxy--galaxy strong lensing systems is strongly dependent on any prior assumptions made about the appearance of the source. Here we present a method of imposing a data-driven prior / regularisation for source galaxies based on denoising diffusion probabilistic models (DDPMs). We use a pre-trained model for galaxy images, AstroDDPM, and a chain of conditional reconstruction steps called denoising diffusion reconstruction model (DDRM) to obtain samples consistent both with the noisy observation and with the distribution of training data for AstroDDPM. We show that these samples hav…
University of Illinois（伊利诺伊大学）；National Center for Supercomputing Applications（国家超级计算应用中心）
Most existing star-galaxy classifiers use the reduced summary information from catalogs, requiring careful feature extraction and selection. The latest advances in machine learning that use deep convolutional neural networks allow a machine to automatically learn the features directly from data, minimizing the need for input from human experts. We present a star-galaxy classification framework that uses deep convolutional neural networks (ConvNets) directly on the reduced, calibrated pixel values. Using data from the Sloan Digital Sky Survey (SDSS) and the Canada-France-Hawaii Telescope Lensin…
The new generation of deep photometric surveys requires unprecedentedly precise shape and photometry measurements of billions of galaxies to achieve their main science goals. At such depths, one major limiting factor is the blending of galaxies due to line-of-sight projection, with an expected fraction of blended galaxies of up to 50%. Current deblending approaches are in most cases either too slow or not accurate enough to reach the level of requirements. This work explores the use of deep neural networks to estimate the photometry of blended pairs of galaxies in monochrome space images, simi…
We discuss the statistical foundations of morphological star-galaxy separation. We show that many of the star-galaxy separation metrics in common use today (e.g. by SDSS or SExtractor) are closely related both to each other, and to the model odds ratio derived in a Bayesian framework by Sebok (1979). While the scaling of these algorithms with the noise properties of the sources varies, these differences do not strongly differentiate their performance. We construct a model of the performance of a star-galaxy separator in a realistic survey to understand the impact of observational signal-to-noi…
We present AstroVaDEr, a variational autoencoder designed to perform unsupervised clustering and synthetic image generation using astronomical imaging catalogues. The model is a convolutional neural network that learns to embed images into a low dimensional latent space, and simultaneously optimises a Gaussian Mixture Model (GMM) on the embedded vectors to cluster the training data. By utilising variational inference, we are able to use the learned GMM as a statistical prior on the latent space to facilitate random sampling and generation of synthetic images. We demonstrate AstroVaDEr's capabi…
Morphologies for DECaLS Galaxies through a combination of non-parametric indices and machine learning methods: A comprehensive catalog using the Galaxy Morphology Extractor (galmex) code
Instituto de Física, Universidad Técnica Federico Santa María（圣塔马利亚联邦理工大学物理研究所）；Millennium Nucleus for Galaxies (MINGAL)（星系千年核（MINGAL））；Departamento de Astronomía, Universidad de La Serena（拉塞雷纳大学天文学系）；Departamento de Astronomía, Facultad Ciencias Físicas y Matemáticas, Universidad de Concepción（康塞普西翁大学理学院天文学系）
The analysis of optical images of galaxy-galaxy strong gravitational lensing systems can provide important information about the distribution of dark matter at small scales. However, the modeling and statistical analysis of these images is extraordinarily complex, bringing together source image and main lens reconstruction, hyper-parameter optimization, and the marginalization over small-scale structure realizations. We present here a new analysis pipeline that tackles these diverse challenges by bringing together many recent machine learning developments in one coherent approach, including va…
We present a new statistical method for constructing background subtracted measurements from event list data gathered by X-ray and gamma ray observatories. This method was initially developed specifically to construct images that account for the high background fraction and low overall count rates observed in survey data from the Mikhail Pavlinsky ART-XC telescope aboard the Spektrum Röntgen Gamma (SRG) mission, although the mathematical underpinnings are valid for data taken with other imaging missions and analysis applications. This method fully accounts for the expected Poisson fluctuations…
Understanding the preferences of transient types for host galaxies with certain characteristics is key to studies of transient physics and galaxy evolution, as well as to transient identification and classification in the LSST era. Here we describe a value-added database of extragalactic transients--supernovae, tidal disruption events, gamma-ray bursts, and other rare events--and their host galaxy properties. Based on reported coordinates, redshifts, and host galaxies (if known) of events, we cross-identify their host galaxies or most likely host candidates in various value-added or survey cat…
We construct a supervised classifier based on Gaussian Mixture Models to probabilistically classify objects in Gaia data release 2 (GDR2) using only photometric and astrometric data in that release. The model is trained empirically to classify objects into three classes -- star, quasar, galaxy -- for G<=14.5 mag down to the Gaia magnitude limit of G=21.0 mag. Galaxies and quasars are identified for the training set by a cross-match to objects with spectroscopic classifications from the Sloan Digital Sky Survey. Stars are defined directly from GDR2. When allowing for the expectation that quasar…
