Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 29: Deep Learning II

T 29.10: Vortrag

Dienstag, 26. März 2019, 18:15–18:30, H06

Image Recognition with Deep Neural Networks for IceAct Air-Cherenkov Telescopes — •Matthias Thiesmeyer, Jan Auffenberg, Pascal Backes, Thomas Bretz, Erik Ganster, Maurice Günder, Merlin Schaufel, Jöran Stettner, and Christopher Wiebusch for the IceCube collaboration — III. Physikalisches Institut B, RWTH Aachen

Deep Neural Networks (DNNs) have brought new possibilities to image analysis. We use these for particle identification in IceAct, which is an array of SiPM-based Imaging Air Cherenkov Telescopes, planned as a surface component of IceCube. One Goal of IceAct is improving composition and gamma ray measurements of the IceCube Neutrino Observatory by the hybrid measurement of air showers. Within the hybrid measurement by the surface detector IceTop, the in-ice detector IceCube, and IceAct, IceTop provides the direction and energy of the shower, IceCube a measurement of the high-energy muon component, and IceAct images the shower development in the atmosphere. We present first results from DNNs trained on simulations of air showers with CORSIKA to separate gamma rays from protons in the energy range from 10-100 TeV.