Deriving star cluster parameters with convolutional neural networks. II. Extinction and cluster/background classification

Abstract

Context. Convolutional neural networks (CNNs) have been established as the go-to method for fast object detection and classification on natural images. This opens the door for astrophysical parameter inference on the exponentially increasing amount of sky survey data. Until now, star cluster analysis was based on integral or resolved stellar photometry, which limits the amount of information that can be extracted from individual pixels of cluster images. Aims. We aim to create a CNN capable of inferring star cluster evolutionary, structural, and environmental parameters from multi-band images, as well to demonstrate its capabilities in discriminating genuine clusters from galactic stellar backgrounds. Methods. A CNN based on the deep residual network (ResNet) architecture was created and trained to infer cluster ages, masses, sizes, and extinctions, with respect to the degeneracies between them. Mock clusters placed on M83 Hubble Space Telescope (HST) images utilizing three photometric passbands (F336W, F438W, and F814W) were used. The CNN is also capable of predicting the likelihood of a cluster's presence in an image, as well as quantifying its visibility (signal-to-noise). Results. The CNN was tested on mock images of artificial clusters and has demonstrated reliable inference results for clusters of ages $\lesssim$100 Myr, extinctions $A_V$ between 0 and 3 mag, masses between $3\times 10^3$ and $3\times 10^5$ ${\mathrm{M}}_{\odot }$, and sizes between 0.04 and 0.4 arcsec at the distance of the M83 galaxy. Real M83 galaxy cluster parameter inference tests were performed with objects taken from previous studies and have demonstrated consistent results.

arXiv comment17 pages, 21 figures

References in corpus (24)

• On the variation of the Initial Mass Function
• PARSEC: stellar tracks and isochrones with the PAdova and TRieste Stellar Evolution Code
• Binary companions of evolved stars in APOGEE DR14: Search method and catalog of ~5,000 companions
• Young massive star clusters
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• The Cepheid Distance to NGC 5236 (M 83) with the VLT
• SLUG -- Stochastically Lighting Up Galaxies. III: A Suite of Tools for Simulated Photometry, Spectroscopy, and Bayesian Inference with Stochastic Stellar Populations
• Accounting for Stochastic Fluctuations when Analysing Integrated Light of Star Clusters. I: First Systematics
• Radio Galaxy Zoo: ClaRAN - A Deep Learning Classifier for Radio Morphologies
• Using H-alpha Morphology and Surface Brightness Fluctuations to Age-Date Star Clusters in M83
• Young Clusters in the Nuclear Starburst of M 83
• Analysing star cluster populations with stochastic models: the HST/WFC3 sample of clusters in M83
• Sizes and Shapes of Young Star Cluster Light Profiles in M83
• Evidence for Environmentally Dependent Cluster Disruption in M83
• An Expanded HST/WFC3 Survey of M83: Project Overview and Targeted Supernova Remnant Search
• Deriving physical parameters of unresolved star clusters. I. Age, mass, and extinction degeneracies
• Big Fish in Small Ponds: Massive Stars in the Low Mass Clusters of M83
• Compact Star Clusters in the M31 Disk
• Deriving physical parameters of unresolved star clusters. II. The degeneracies of age, mass, extinction, and metallicity
• The First Metallicity Study of M83 using the integrated UV light of Star Clusters
• Deriving star cluster parameters with convolutional neural networks. I. Age, mass, and size
• Galaxy shape measurement with convolutional neural networks
• Deriving physical parameters of unresolved star clusters IV. The M33 star cluster system
• Hierarchical Bayesian approach for estimating physical properties in nearby galaxies: Age Maps (Paper II)

Cited by in corpus (2)

• Study of star clusters in the M83 galaxy with a convolutional neural network
• Predicting Images for the Dynamics Of stellar Clusters (π-DOC): a deep learning framework to predict mass, distance and age of globular clusters