Karlsruhe 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 119: Data, AI, Computing 8 (foundational & transformer models)
T 119.4: Talk
Friday, March 8, 2024, 09:45–10:00, Geb. 30.33: MTI
Point-Clouds based Diffusion Model on Hadronic Shower — •Martina Mozzanica1, Erik Buhmann1, Frank Gaede2,3, Gregor Kasieczka1,3, Anatolii Korol2, William Korcari1, 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
Simulating showers of particles in highly-granular detectors is a key frontier in the application of machine learning to particle physics. Achieving high accuracy and speed with generative machine learning models can enable them to augment traditional simulations and alleviate a major computing constraint.
Recent developments have shown how diffusion based generative shower simulation approach that do not rely on a fixed structure, but instead generates geometry-independent point clouds are very efficient. We present an extension to a point-cloud based diffusion model, i.e. CaloClouds, previously applied only to electromagnetic showers of the International Large Calorimeter (ILD).
The works focuses on the more challenging hadronic showers, namely pion showers, and introduces a more advanced architecture that successfully deals with the increasing complexity of the data, i.e. the attention mechanism.
Keywords: generative models; diffusion models; hadronic showers; calorimeters; machine learning