Seminarios de Instrumentación
Deep Learning for Full-Event Reconstruction in Imaging Atmospheric Cherenkov Telescopes: The CTLearn Package

Abstract

Gamma rays are key when it comes to probing a wide range of astrophysical phenomena that provide a deeper understanding of the most energetic events in the universe, as well as topics such as dark matter searches from an indirect perspective or the Lorentz invariance. They can be detected through Imaging Atmospheric Cherenkov Telescopes (IACTs), which capture images of extensive air showers generated by gamma rays and cosmic rays when they interact with the atmosphere. One of the main challenges about this data is the reconstruction of the event's properties, i.e., estimating the direction of arrival, energy, and type (gamma ray, proton, electron, etc.) of the particles that triggered the shower. AI techniques, such as deep learning methods, have been demonstrated to be suitable for the reconstruction of these events since they are used to analyze and exploit loads of data for carrying out classification and characterization tasks.