A Generative Model of Galactic Dust Emission Using Variational Inference

Abstract

Emission from the interstellar medium can be a significant contaminant of measurements of the intensity and polarization of the cosmic microwave background (CMB). For planning CMB observations, and for optimizing foreground-cleaning algorithms, a description of the statistical properties of such emission can be helpful. Here we examine a machine learning approach to inferring the statistical properties of dust from either observational data or physics-based simulations. In particular, we apply a type of neural network called a Variational Auto Encoder (VAE) to maps of the intensity of emission from interstellar dust as inferred from Planck sky maps and demonstrate its ability to a) simulate new samples with similar summary statistics as the training set, b) provide fits to emission maps withheld from the training set, and c) produce constrained realizations. We find VAEs are easier to train than another popular architecture: that of Generative Adversarial Networks (GANs), and are better-suited for use in Bayesian inference.

10 pages, 6 figures, 2 tables

References in corpus (33)

• WaveNet: A Generative Model for Raw Audio
• Large Scale GAN Training for High Fidelity Natural Image Synthesis
• An Introduction to Variational Autoencoders
• Stochastic Backpropagation and Approximate Inference in Deep Generative Models
• Pixel Recurrent Neural Networks
• The Simons Observatory: Science goals and forecasts
• NICE: Non-linear Independent Components Estimation
• Signature of Gravity Waves in Polarization of the Microwave Background
• A Probe of Primordial Gravity Waves and Vorticity
• MASTER of the CMB Anisotropy Power Spectrum: A Fast Method for Statistical Analysis of Large and Complex CMB Data Sets
• Density estimation using Real NVP
• BICEP2 / Keck Array x: Constraints on Primordial Gravitational Waves using Planck, WMAP, and New BICEP2/Keck Observations through the 2015 Season
• A unified pseudo-$C_{\ell }$ framework
• SPT-3G: A Next-Generation Cosmic Microwave Background Polarization Experiment on the South Pole Telescope
• DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications
• Improving Variational Inference with Inverse Autoregressive Flow
• The Pseudo-$C_l$ method: Cosmic microwave background anisotropy power spectrum statistics for high precision cosmology
• Cosmic microwave background anisotropy power spectrum statistics for high precision cosmology
• On Determining the Spectrum of Primordial Inhomogeneity from the COBE DMR Sky Maps: I. Method
• Towards Deeper Understanding of Variational Autoencoding Models
• A brute force analysis of the COBE DMR data
• Generating Diverse High-Fidelity Images with VQ-VAE-2
• BICEP Array: a multi-frequency degree-scale CMB polarimeter
• Detectability of Tensor Perturbations Through CBR Anisotropy (final published version)
• Bayesian delensing delight: sampling-based inference of the primordial CMB and gravitational lensing
• Full-sky Cosmic Microwave Background Foreground Cleaning Using Machine Learning
• The Atacama Cosmology Telescope: Two-season ACTPol Extragalactic Point Sources and their Polarization properties
• Convolutional Neural Networks on the HEALPix sphere: a pixel-based algorithm and its application to CMB data analysis
• Filling in CMB map missing data using constrained Gaussian realizations
• Dust Polarization Maps from TIGRESS: E/B power asymmetry and TE correlation
• Using inpainting to construct accurate cut-sky CMB estimators
• Cleaning our own Dust: Simulating and Separating Galactic Dust Foregrounds with Neural Networks
• Uncertainty Quantification with Generative Models

Cited by in corpus (11)

• A new approach for the statistical denoising of Planck interstellar dust polarization data
• Non-Gaussian modelling and statistical denoising of Planck dust polarization full-sky maps using scattering transforms
• MillimeterDL: Deep Learning Simulations of the Microwave Sky
• A Robust Study of High-Redshift Galaxies: Unsupervised Machine Learning for Characterising morphology with JWST up to z ~ 8
• Full-sky, arcminute-scale, 3D models of Galactic microwave foreground dust emission based on filaments
• Generative Models of Multi-channel Data from a Single Example -- Application to Dust Emission
• Normalizing flows for random fields in cosmology
• PCAT-DE: Reconstructing point-like and diffuse signals in astronomical images using spatial and spectral information
• Deep Needlet: A CNN based full sky component separation method in Needlet space
• Euclid preparation. LXVII. Deep learning true galaxy morphologies for weak lensing shear bias calibration
• ForSE+: Simulating non-Gaussian CMB foregrounds at 3 arcminutes in a stochastic way based on a generative adversarial network