Neural Posterior Estimation for Cataloging Astronomical Images with Spatially Varying Backgrounds and Point Spread Functions

Abstract

Neural posterior estimation (NPE), a type of amortized variational inference, is a computationally efficient means of constructing probabilistic catalogs of light sources from astronomical images. To date, NPE has not been used to perform inference in models with spatially varying covariates. However, ground-based astronomical images have spatially varying sky backgrounds and point spread functions (PSFs), and accounting for this variation is essential for constructing accurate catalogs of imaged light sources. In this work, we introduce a method of performing NPE with spatially varying backgrounds and PSFs. In this method, we generate synthetic catalogs and semi-synthetic images for these catalogs using randomly sampled PSF and background estimates from existing surveys. Using this data, we train a neural network, which takes an astronomical image and representations of its background and PSF as input, to output a probabilistic catalog. Our experiments with Sloan Digital Sky Survey data demonstrate the effectiveness of NPE in the presence of spatially varying backgrounds and PSFs for light source detection, star/galaxy separation, and flux measurement.

arXiv commentPublished in the Astronomical Journal

References in corpus (32)

• Domain Generalization: A Survey
• Improved background subtraction for the Sloan Digital Sky Survey images
• Robust Machine Learning Applied to Astronomical Datasets I: Star-Galaxy Classification of the SDSS DR3 Using Decision Trees
• SCARLET: Source separation in multi-band images by Constrained Matrix Factorization
• The LSST DESC DC2 Simulated Sky Survey
• The three-year shear catalog of the Subaru Hyper Suprime-Cam SSP Survey
• Dark Energy Survey Year 3 Results: Point-Spread Function Modeling
• Interpolating point spread function anisotropy
• Removing Internal Reflections from Deep Imaging Datasets
• Identifying galaxies, quasars, and stars with machine learning: A new catalogue of classifications for 111 million SDSS sources without spectra
• Simulation of Astronomical Images from Optical Survey Telescopes using a Comprehensive Photon Monte Carlo Approach
• Deep Generative Models for Galaxy Image Simulations
• Star-galaxy classification in the Dark Energy Survey Y1 dataset
• Probabilistic Catalogs for Crowded Stellar Fields
• The effect of FIR emission from SDSS galaxies on the SFD Galactic extinction map
• Robust Simulation-Based Inference in Cosmology with Bayesian Neural Networks
• Improved Point Source Detection in Crowded Fields using Probabilistic Cataloging
• Point Spread Function Modelling for Wide Field Small Aperture Telescopes with a Denoising Autoencoder
• Cosmic shear requirements on the wavelength-dependence of telescope point spread functions
• A method of complex background estimation in astronomical images
• A Study of the Point Spread Function in SDSS Images
• Multiband Probabilistic Cataloging: A Joint Fitting Approach to Point Source Detection and Deblending
• A Recipe for Unbiased Background Modeling in Deep Wide-Field Astronomical Images
• Learning Robust Statistics for Simulation-based Inference under Model Misspecification
• Robust Neural Posterior Estimation and Statistical Model Criticism
• Completeness - I. Revised, reviewed and revived
• Variational Inference for Deblending Crowded Starfields
• HNPE: Leveraging Global Parameters for Neural Posterior Estimation
• PCAT-DE: Reconstructing point-like and diffuse signals in astronomical images using spatial and spectral information
• nbi: the Astronomer's Package for Neural Posterior Estimation
• Scalable Bayesian Inference for Detection and Deblending in Astronomical Images
• Globally Convergent Variational Inference