SMuK 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 10: ML Methods I
T 10.3: Vortrag
Montag, 20. März 2023, 17:00–17:15, HSZ/0405
DeepTreeGAN: Fast Generation of High Dimensional Point Clouds for Calorimeter Simulation — •Moritz Scham1,2,3, Dirk Krücker1, and Kerstin Borras1,2 — 1Deutsches Elektronen-Synchrotron, Hamburg, Germany — 2RWTH Aachen University - III. Physikalisches Institut A, Aachen, Germany — 3Institute for Advanced Simulation - Jülich Supercomputing Centre, Juelich, Germany
In high energy physics, detailed and time-consuming simulations are used for particle interactions with detectors. To bypass these simulations with a generative model, the generation of large point clouds in a short time is required, while the complex dependencies between the particles must be correctly modeled. Particle showers are inherently tree-based processes, as each particle is produced by decays or detector interaction of a particle of the previous generation.
In this work, we present a novel GNN model that is able to generate such point clouds in a tree-based manner. We show that this model is able to reproduce complex distributions, and we evaluate its performance on the public JetNet Dataset.