Identification of point sources in gamma rays using U-shaped convolutional neural networks and a data challenge

Abstract

At GeV energies, the sky is dominated by the interstellar emission from the Galaxy. With limited statistics and spatial resolution, accurately separating point sources is therefore challenging. Here we present the first application of deep learning based algorithms to automatically detect and classify point sources from gamma-ray data. For concreteness we refer to this approach as AutoSourceID. To detect point sources, we utilized U-shaped convolutional networks for image segmentation and {\it k}-means for source clustering and localization. We also explored the Centroid-Net algorithm, which is designed to find and count objects. The training data are based on 9.5 years of exposure from The Fermi Large Area Telescope (Fermi-LAT) and we used source properties of active galactic nuclei (AGNs) and pulsars (PSRs) from the fourth Fermi-LAT source catalog (4FGL) in addition to several models of background interstellar emission. The results of the localization algorithm are fed into a classification neural network that is trained to separate the three general source classes (AGNs, PSRs, and FAKE sources). We compared our localization algorithms qualitatively with traditional methods and find them to have similar detection thresholds. We also demonstrate the robustness of our source localization algorithms to modifications in the interstellar emission models, which presents a clear advantage over traditional methods. The classification network is able to discriminate between the three classes with typical accuracy of $\sim$ 70%, as long as balanced data sets are used in classification training. In https://github.com/bapanes/AutoSourceID, we publish our training data sets and analysis scripts and invite the community to join the data challenge aimed to improve the localization and classification of gamma-ray point sources.

19 pages (two-column format), 14 figures, accepted for publication in Astronomy & Astrophysics

References in corpus (32)

• Adam: A Method for Stochastic Optimization
• HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
• Asymptotic formulae for likelihood-based tests of new physics
• Deep Learning using Rectified Linear Units (ReLU)
• Fermi Large Area Telescope Fourth Source Catalog
• The Second Fermi Large Area Telescope Catalog of Gamma-ray Pulsars
• Dark Matter Annihilation in The Galactic Center As Seen by the Fermi Gamma Ray Space Telescope
• Background model systematics for the Fermi GeV excess
• Development of the Model of Galactic Interstellar Emission for Standard Point-Source Analysis of Fermi Large Area Telescope Data
• The Fourth Catalog of Active Galactic Nuclei Detected by the Fermi Large Area Telescope
• Robust Lane Detection from Continuous Driving Scenes Using Deep Neural Networks
• The 1st Fermi Lat Supernova Remnant Catalog
• Machine Learning for Precipitation Nowcasting from Radar Images
• Lunar Crater Identification via Deep Learning
• Learning to Predict the Cosmological Structure Formation
• The Consistency of Fermi-LAT Observations of the Galactic Center with a Millisecond Pulsar Population in the Central Stellar Cluster
• Cosmic Ray Protons in the Inner Galaxy and the Galactic Center Gamma-Ray Excess
• Pulsars Cannot Account for the Inner Galaxy's GeV Excess
• Galactic Center gamma-ray "excess" from an active past of the Galactic Centre?
• Millisecond pulsars and the Galactic Center gamma-ray excess: the importance of luminosity function and secondary emission
• The Fermi-LAT GeV Excess Traces Stellar Mass in the Galactic Bulge
• Classification and Ranking of Fermi LAT Gamma-ray Sources from the 3FGL Catalog using Machine Learning Techniques
• Discovery of two millisecond pulsars in Fermi sources with the Nancay Radio Telescope
• Dark matter vs. Pulsars: Catching the impostor
• DeepCMB: Lensing Reconstruction of the Cosmic Microwave Background with Deep Neural Networks
• Blazar Flaring Patterns (B-FlaP): Classifying Blazar Candidates of Uncertain type in the third Fermi-LAT catalog by Artificial Neural Networks
• DeepSource: Point Source Detection using Deep Learning
• Deriving the contribution of blazars to the Fermi-LAT Extragalactic $\gamma$-ray background at $E>10$ GeV with efficiency corrections and photon statistics
• Unidentified Gamma-ray Sources as Targets for Indirect Dark Matter Detection with the Fermi-Large Area Telescope
• Analyzing γ-rays of the Galactic Center with Deep Learning
• Blazar -- IceCube neutrino association revisited
• Harnessing the Population Statistics of Subhalos to Search for Annihilating Dark Matter

Cited by in corpus (7)

• The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider
• A neural simulation-based inference approach for characterizing the Galactic Center $\gamma$-ray excess
• Machine learning methods for constructing probabilistic Fermi-LAT catalogs
• AutoSourceID-Light. Fast Optical Source Localization via U-Net and Laplacian of Gaussian
• Searching for dark matter subhalos in the Fermi-LAT catalog with Bayesian neural networks
• AutoSourceID-FeatureExtractor. Optical image analysis using a two-step mean variance estimation network for feature estimation and uncertainty characterisation
• Classification of Fermi-LAT blazars with Bayesian neural networks