CENN: a fully convolutional neural network for CMB recovery in realistic microwave sky simulations

Abstract

Context. Component separation is the process to extract the sources of emission in astrophysical maps generally by taking into account multi-frequency information. Developing more reliable methods to perform component separation is crucial for future cosmic microwave background (CMB) experiments such as the Simons Observatory, the CMB-S4 or the LiteBIRD satellite. Aims. We aim to develop a machine learning method based on fully convolutional neural networks called the Cosmic microwave background Extraction Neural Network (CENN) in order to extract the CMB signal in total intensity by training the network with realistic simulations. The frequencies used are the Planck channels 143, 217 and 353 GHz, and we validate the neural network at all sky, and at three latitude intervals: 0Moreover, we do not use any Galactic or point source masks. Methods. To train the neural network, we produce multi-frequency realistic simulations in the form of patches of 256 × 256 pixels, which contain the CMB signal, the Galactic thermal dust, CIB and PS emissions, the thermal Sunyaev-Zel'dovich effect from galaxy clusters and the instrumental noise. After validate the network, we compare the power spectrum from input and output maps. We analyse the power spectrum from the residuals at each latitude interval and at all sky and we study the performance of our model dealing with high contamination at small scales. Results. We obtain a CMB power spectrum with a mean error of 13 ± 113 µK 2 for multipoles up to above 4 000. We compute the residuals for all sky and for each latitude interval, obtaining 7 ± 25 µK 2 for 0 • < |b| < 5 • , 2 ± 10 µK 2 for 5 • < |b| < 30 • and 2 ± 3 µK 2 for 30 • < |b| < 90 • for multipoles up to above 4 000. For all sky, we obtain residuals of 0.2 ± 0.5 µK 2 for l <= 1 000 and 5 ± 12 µK 2 for l <= 4 000. We validate the neural network in a single patch with strong contamination at small scales, obtaining an error of 50 ± 120 µK 2 and residuals of 40 ± 10 µK 2 up to l ∼ 2 500. Conclusions. Based on the results, fully convolutional neural networks are promising methods to perform component separation in future CMB experiments. Moreover, we show that CENN is reliable against different levels of contamination from Galactic and point source foregrounds at both large and small scales.

Cosmic Microwave Background Recovery: A Graph-Based Bayesian Convolutional Network Approach

“…As previously noted, the model underestimates regions where the CMB temperature has a high magnitude and predicts high uncertainty in these areas. We find this issue in related works (Petroff et al 2020;Wang et al 2022;Casas et al 2022) as well, where there are structured differences in the true and recovered spectra, which cannot be accounted for by randomness or noise. Resolving this issue will improve prediction accuracy and uncertainty quantification calibration.…”

“…CNNs retain spatial correlations but are formulated for Euclidean input. There has been progress in training CNNs using flattened 2D patches of full and partial-sky maps (Wang et al 2022;Casas et al 2022); however, these 2D projections are prone to edge artifacts. DeepSphere (Perraudin et al 2019) presents an efficient spherical CNN that leverages the hierarchical HEALPix (Gorski et al 2005) graph representations for pooling and performs convolution in the spectral domain.…”

Relationship of Neural Correlates of Gait Characteristics and Cognitive Dysfunction in Patients With Mild Cognitive Impairment

“…CMB data encapsulates the thermal radiation left over from the Big Bang, the origin of the universe. Machine learning has been applied to predict the posterior distribution of the cosmological parameters (Hortúa et al, 2020) as well as signal recovery and dust cleaning (Caldeira et al, 2019;Wang et al, 2022;Casas et al, 2022). A potential application of machine learning for CMB data is to increase resolution of the data, which has mainly been progressed from the physical development of new sensors (Bennett et al, 1996;Spergel et al, 2003;Ade et al, 2014).…”

Likelihood-free Inference with the Mixture Density Network

“…Recently, the artificial neural network (ANN) with many hidden layers has achieved significant improvement in many fields including cosmology and astrophysics. For example, it performs excellently in searching for and estimating the parameters of strong gravitational lenses (Hezaveh et al 2017;Jacobs et al 2017;Petrillo et al 2017;Pourrahmani et al 2018;Schaefer et al 2018;Li et al , 2021, analyzing gravitational waves (George et al 2018a;George & Huerta 2018b, 2018cShen et al 2019;, classifying the large-scale structure of the universe (Aragon-Calvo 2019), discriminating between cosmological and reionization models (Schmelzle et al 2017;Hassan et al 2018), recovering the cosmic microwave background (CMB) signal from contaminated observations (Petroff et al 2020;Casas et al 2022;, reconstructing functions from observational data (Wang et al 2020b;Escamilla-Rivera et al 2020;Wang et al 2021), and even estimating cosmological and astrophysical parameters (Shimabukuro & Semelin 2017;Fluri et al 2018;Schmit & Pritchard 2018;Fluri et al 2019;Ribli et al 2019;Wang et al 2020a;Ntampaka et al 2020;Nygaard et al 2022).…”