Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 120: Data, AI, Computing 9 (generative models & simulation)
T 120.4: Vortrag
Freitag, 8. März 2024, 09:45–10:00, Geb. 30.34: LTI
Equivariant Point Cloud GAN for 4-dimensional Calorimeter Clouds. — •william korcari1, erik buhmann1, frank gaede2, gregor kasieczka1,3, anatolii korol2, katja krüger2, and peter mckeown2 — 1Institut für Experimentalphysik, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany — 2Deutsches Elektronen-Synchrotron DESY, Notkestr. 85, 22607 Hamburg, Germany — 3Center for Data and Computing in Natural Sciences CDCS, Deutsches Elektronen-Synchrotron DESY, Notkestr. 85, 22607 Hamburg, Germany
Fast simulation of the energy depositions in high-granular detectors is crucial for future collider experiments with increasing luminosity. Many proofs of concepts show how generative machine learning models can speed up and augment the traditional simulation chain in physics analysis. EPiC GAN has already shown promising results with very high-fidelity simulation of the physics of top jets with cardinality going as high as 150 particles and characterized by 3 dimensions. We show an extension of such a model, capable of conditional generation of photon calorimeter showers with even higher cardinality and an extra dimension.
Keywords: generative model; GAN; calorimeter; showers; machine learning