First IACT waveform analysis based on deep convolutional neural networks using CTLearn

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985815%3A_____%2F25%3A00646040" target="_blank" >RIV/67985815:_____/25:00646040 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/68378271:_____/25:00646040 RIV/61989592:15310/25:73635264

Výsledek na webu

<a href="https://pos.sissa.it/501/758/pdf" target="_blank" >https://pos.sissa.it/501/758/pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.22323/1.501.0758" target="_blank" >10.22323/1.501.0758</a>

Jazyk výsledku

angličtina

Název v původním jazyce

First IACT waveform analysis based on deep convolutional neural networks using CTLearn

Popis výsledku v původním jazyce

Imaging atmospheric Cherenkov telescopes (IACTs) detect extended air showers (EASs) generated when very-high-energy (VHE) gamma rays or cosmic rays interact with the Earth's atmosphere. Cherenkov photons produced during an EAS are captured by fast-imaging cameras, which record both the spatial and temporal development of the shower, along with calorimetric data. By analyzing these recordings, the properties of the original VHE particle—such as its type, energy, and direction of arrival—can be reconstructed through machine learning techniques. This contribution focuses on the Large-Sized Telescopes (LSTs) of the Cherenkov Telescope Array Observatory, a next-generation ground-based gamma-ray observatory. LSTs are responsible for reconstructing lower-energy gamma rays in the tens of GeV range. We explore a novel event reconstruction technique based on deep convolutional neural networks (CNNs) applied on calibrated and cleaned waveforms of the IACT camera pixels using CTLearn. Our approach explicitly incorporates the time development of the shower, enabling a more accurate reconstruction of the event. This method eliminates the need for charge integration or handcrafted feature extraction, allowing the model to directly learn from waveform data.

Název v anglickém jazyce

First IACT waveform analysis based on deep convolutional neural networks using CTLearn

Popis výsledku anglicky

Imaging atmospheric Cherenkov telescopes (IACTs) detect extended air showers (EASs) generated when very-high-energy (VHE) gamma rays or cosmic rays interact with the Earth's atmosphere. Cherenkov photons produced during an EAS are captured by fast-imaging cameras, which record both the spatial and temporal development of the shower, along with calorimetric data. By analyzing these recordings, the properties of the original VHE particle—such as its type, energy, and direction of arrival—can be reconstructed through machine learning techniques. This contribution focuses on the Large-Sized Telescopes (LSTs) of the Cherenkov Telescope Array Observatory, a next-generation ground-based gamma-ray observatory. LSTs are responsible for reconstructing lower-energy gamma rays in the tens of GeV range. We explore a novel event reconstruction technique based on deep convolutional neural networks (CNNs) applied on calibrated and cleaned waveforms of the IACT camera pixels using CTLearn. Our approach explicitly incorporates the time development of the shower, enabling a more accurate reconstruction of the event. This method eliminates the need for charge integration or handcrafted feature extraction, allowing the model to directly learn from waveform data.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10308 - Astronomy (including astrophysics,space science)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of Science

ISBN

—

ISSN

1824-8039

e-ISSN

—

Počet stran výsledku

10

Strana od-do

758

Název nakladatele

Sissa Medilab srl

Místo vydání

Trieste

Místo konání akce

Ženeva

Datum konání akce

15. 7. 2025

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—