Records |
Author |
Panes, B.; Eckner, C.; Hendriks, L.; Caron, S.; Dijkstra, K.; Johannesson, G.; Ruiz de Austri, R.; Zaharijas, G. |
Title |
Identification of point sources in gamma rays using U-shaped convolutional neural networks and a data challenge |
Type |
Journal Article |
Year |
2021 |
Publication |
Astronomy & Astrophysics |
Abbreviated Journal |
Astron. Astrophys. |
Volume |
656 |
Issue |
|
Pages |
A62 - 18pp |
Keywords |
catalogs; gamma rays: general; astroparticle physics; methods: numerical; methods: data analysis; techniques: image processing |
Abstract |
Context. 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. Aims. 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. Methods. To detect point sources, we utilized U-shaped convolutional networks for image segmentation and k-means for source clustering and localization. We also explored the Centroid-Net algorithm, which is designed to find and count objects. Using two algorithms allows for a cross check of the results, while a combination of their results can be used to improve performance. 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 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). Results. 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 similar to 70%, as long as balanced data sets are used in classification training. We published online 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. |
Address |
[Panes, Boris] Pontificia Univ Catolica Chile, Ave Vicuna Mackenna 4860, Macul, Region Metropol, Chile, Email: bapanes@gmail.com |
Corporate Author |
|
Thesis |
|
Publisher |
Edp Sciences S A |
Place of Publication |
|
Editor |
|
Language |
English |
Summary Language |
|
Original Title |
|
Series Editor |
|
Series Title |
|
Abbreviated Series Title |
|
Series Volume |
|
Series Issue |
|
Edition |
|
ISSN |
0004-6361 |
ISBN |
|
Medium |
|
Area |
|
Expedition |
|
Conference |
|
Notes |
WOS:000725877600001 |
Approved |
no |
Is ISI |
yes |
International Collaboration |
yes |
Call Number |
IFIC @ pastor @ |
Serial |
5053 |
Permanent link to this record |
|
|
|
Author |
Pierre Auger Collaboration (Aab, A. et al); Pastor, S. |
Title |
A search for point sources of EeV photons |
Type |
Journal Article |
Year |
2014 |
Publication |
Astrophysical Journal |
Abbreviated Journal |
Astrophys. J. |
Volume |
789 |
Issue |
2 |
Pages |
160 - 12pp |
Keywords |
astroparticle physics; cosmic rays; methods: data analysis |
Abstract |
Measurements of air showers made using the hybrid technique developed with the fluorescence and surface detectors of the Pierre Auger Observatory allow a sensitive search for point sources of EeV photons anywhere in the exposed sky. A multivariate analysis reduces the background of hadronic cosmic rays. The search is sensitive to a declination band from -85 degrees to +20 degrees, in an energy range from 10(17.3) eV to 10(18.5) eV. No photon point source has been detected. An upper limit on the photon flux has been derived for every direction. The mean value of the energy flux limit that results from this, assuming a photon spectral index of -2, is 0.06 eV cm(-2) s(-1), and no celestial direction exceeds 0.25 eV cm(-2) s(-1). These upper limits constrain scenarios in which EeV cosmic ray protons are emitted by non-transient sources in the Galaxy. |
Address |
[Aab, A.; Buchholz, P.; Erfani, M.; Froehlich, U.; Heimann, P.; Homola, P.; Kuempel, D.; Niechciol, M.; Ochilo, L.; Risse, M.; Yushkov, A.; Ziolkowski, M.] Univ Siegen, D-57068 Siegen, Germany |
Corporate Author |
|
Thesis |
|
Publisher |
Iop Publishing Ltd |
Place of Publication |
|
Editor |
|
Language |
English |
Summary Language |
|
Original Title |
|
Series Editor |
|
Series Title |
|
Abbreviated Series Title |
|
Series Volume |
|
Series Issue |
|
Edition |
|
ISSN |
0004-637x |
ISBN |
|
Medium |
|
Area |
|
Expedition |
|
Conference |
|
Notes |
WOS:000338674900069 |
Approved |
no |
Is ISI |
yes |
International Collaboration |
yes |
Call Number |
IFIC @ pastor @ |
Serial |
1842 |
Permanent link to this record |
|
|
|
Author |
Pierre Auger Collaboration (Aab, A. et al); Pastor, S. |
Title |
A targeted search for point sources of EeV neutrons |
Type |
Journal Article |
Year |
2014 |
Publication |
Astrophysical Journal Letters |
Abbreviated Journal |
Astrophys. J. Lett. |
Volume |
789 |
Issue |
2 |
Pages |
L34 - 7pp |
Keywords |
cosmic rays; Galaxy: disk; methods: data analysis |
Abstract |
A flux of neutrons from an astrophysical source in the Galaxy can be detected in the Pierre Auger Observatory as an excess of cosmic-ray air showers arriving from the direction of the source. To avoid the statistical penalty for making many trials, classes of objects are tested in combinations as nine “target sets,” in addition to the search for a neutron flux from the Galactic center or from the Galactic plane. Within a target set, each candidate source is weighted in proportion to its electromagnetic flux, its exposure to the Auger Observatory, and its flux attenuation factor due to neutron decay. These searches do not find evidence for a neutron flux from any class of candidate sources. Tabulated results give the combined p-value for each class, with and without the weights, and also the flux upper limit for the most significant candidate source within each class. These limits on fluxes of neutrons significantly constrain models of EeV proton emission from non-transient discrete sources in the Galaxy. |
Address |
[Aab, A.; Buchholz, P.; Erfani, M.; Frohlich, U.; Heimann, P.; Homola, P.; Niechciol, M.; Ochilo, L.; Risse, M.; Yushkov, A.; Ziolkowski, M.] Univ Siegen, D-57068 Siegen, Germany |
Corporate Author |
|
Thesis |
|
Publisher |
Iop Publishing Ltd |
Place of Publication |
|
Editor |
|
Language |
English |
Summary Language |
|
Original Title |
|
Series Editor |
|
Series Title |
|
Abbreviated Series Title |
|
Series Volume |
|
Series Issue |
|
Edition |
|
ISSN |
2041-8205 |
ISBN |
|
Medium |
|
Area |
|
Expedition |
|
Conference |
|
Notes |
WOS:000339876800009 |
Approved |
no |
Is ISI |
yes |
International Collaboration |
yes |
Call Number |
IFIC @ pastor @ |
Serial |
1885 |
Permanent link to this record |
|
|
|
Author |
Pierre Auger Collaboration (Abreu, P. et al); Pastor, S. |
Title |
A search for point sources of EeV neutrons |
Type |
Journal Article |
Year |
2012 |
Publication |
Astrophysical Journal |
Abbreviated Journal |
Astrophys. J. |
Volume |
760 |
Issue |
2 |
Pages |
148 - 11pp |
Keywords |
cosmic rays; Galaxy: disk; methods: data analysis |
Abstract |
A thorough search of the sky exposed at the Pierre Auger Cosmic Ray Observatory reveals no statistically significant excess of events in any small solid angle that would be indicative of a flux of neutral particles from a discrete source. The search covers from -90 degrees to +15 degrees in declination using four different energy ranges above 1 EeV (10(18) eV). The method used in this search is more sensitive to neutrons than to photons. The upper limit on a neutron flux is derived for a dense grid of directions for each of the four energy ranges. These results constrain scenarios for the production of ultrahigh energy cosmic rays in the Galaxy. |
Address |
[Abreu, P.; Andringa, S.; Assis, P.; Brogueira, P.; Cazon, L.; Conceicao, R.; Diogo, F.; Espadanal, J.; Goncalves, P.; Pimenta, M.; Santo, C. E.; Santos, E.; Tome, B.] Univ Tecn Lisboa, LIP, Lisbon, Portugal |
Corporate Author |
|
Thesis |
|
Publisher |
Iop Publishing Ltd |
Place of Publication |
|
Editor |
|
Language |
English |
Summary Language |
|
Original Title |
|
Series Editor |
|
Series Title |
|
Abbreviated Series Title |
|
Series Volume |
|
Series Issue |
|
Edition |
|
ISSN |
0004-637x |
ISBN |
|
Medium |
|
Area |
|
Expedition |
|
Conference |
|
Notes |
WOS:000311217000052 |
Approved |
no |
Is ISI |
yes |
International Collaboration |
yes |
Call Number |
IFIC @ pastor @ |
Serial |
1218 |
Permanent link to this record |
|
|
|
Author |
Stoppa, F.; Bhattacharyya, S.; Ruiz de Austri, R.; Vreeswijk, P.; Caron, S.; Zaharijas, G.; Bloemen, S.; Principe, G.; Malyshev, D.; Vodeb, V.; Groot, P.J.; Cator, E.; Nelemans, G. |
Title |
AutoSourceID-Classifier Star-galaxy classification using a convolutional neural network with spatial information |
Type |
Journal Article |
Year |
2023 |
Publication |
Astronomy & Astrophysics |
Abbreviated Journal |
Astron. Astrophys. |
Volume |
680 |
Issue |
|
Pages |
A109 - 16pp |
Keywords |
methods: data analysis; techniques: image processing; astronomical databases: miscellaneous; stars: imaging; Galaxies: statistics |
Abstract |
Aims. Traditional star-galaxy classification techniques often rely on feature estimation from catalogs, a process susceptible to introducing inaccuracies, thereby potentially jeopardizing the classification's reliability. Certain galaxies, especially those not manifesting as extended sources, can be misclassified when their shape parameters and flux solely drive the inference. We aim to create a robust and accurate classification network for identifying stars and galaxies directly from astronomical images.Methods. The AutoSourceID-Classifier (ASID-C) algorithm developed for this work uses 32x32 pixel single filter band source cutouts generated by the previously developed AutoSourceID-Light (ASID-L) code. By leveraging convolutional neural networks (CNN) and additional information about the source position within the full-field image, ASID-C aims to accurately classify all stars and galaxies within a survey. Subsequently, we employed a modified Platt scaling calibration for the output of the CNN, ensuring that the derived probabilities were effectively calibrated, delivering precise and reliable results.Results. We show that ASID-C, trained on MeerLICHT telescope images and using the Dark Energy Camera Legacy Survey (DECaLS) morphological classification, is a robust classifier and outperforms similar codes such as SourceExtractor. To facilitate a rigorous comparison, we also trained an eXtreme Gradient Boosting (XGBoost) model on tabular features extracted by SourceExtractor. While this XGBoost model approaches ASID-C in performance metrics, it does not offer the computational efficiency and reduced error propagation inherent in ASID-C's direct image-based classification approach. ASID-C excels in low signal-to-noise ratio and crowded scenarios, potentially aiding in transient host identification and advancing deep-sky astronomy. |
Address |
[Stoppa, F.; Vreeswijk, P.; Bloemen, S.; Groot, P. J.; Nelemans, G.] Radboud Univ Nijmegen, Dept Astrophys IMAPP, POB 9010, NL-6500 GL Nijmegen, Netherlands, Email: f.stoppa@astro.ru.nl |
Corporate Author |
|
Thesis |
|
Publisher |
Edp Sciences S A |
Place of Publication |
|
Editor |
|
Language |
English |
Summary Language |
|
Original Title |
|
Series Editor |
|
Series Title |
|
Abbreviated Series Title |
|
Series Volume |
|
Series Issue |
|
Edition |
|
ISSN |
0004-6361 |
ISBN |
|
Medium |
|
Area |
|
Expedition |
|
Conference |
|
Notes |
WOS:001131898100001 |
Approved |
no |
Is ISI |
yes |
International Collaboration |
yes |
Call Number |
IFIC @ pastor @ |
Serial |
5888 |
Permanent link to this record |