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

T: Fachverband Teilchenphysik

T 43: Data, AI, Computing 3 (pointclouds & graphs)

T 43.3: Vortrag

Dienstag, 5. März 2024, 16:30–16:45, Geb. 30.33: MTI

Graph Neural Network based Hit Classification for Tracking at Belle II — •Greta Heine, Torben Ferber, Lea Reuter, and Slavomira Stefkova — Institute of Experimental Particle Physics (ETP), Karlsruhe Institute of Technology (KIT)

Over the next few years, the Belle II Experiment will increase its instantaneous luminosity, which will also lead to a significant increase in the beam background, affecting the efficiency of both online and offline tracking algorithms. To overcome this challenge and to facilitate the identification of displaced vertices for the discovery of new physics phenomena, Belle II needs a more robust tracking algorithm on trigger level.

Graph Neural Networks (GNNs), with their ability to model complex relationships within detector hits, are well suited for tracking and are currently under investigation by Belle II, particularly in the context of object condensation for the Belle II Central Drift Chamber. Due to strict timing constraints, especially in the real-time application in the hardware trigger system, it becomes imperative to clean-up the detector hits from background noise. This talk presents first studies on hit clean-up in the context of anticipated high beam background conditions of the Belle II Experiment based on GNN edge classification using detector-level information.

Keywords: GNN; Tracking; Belle II; FPGA; Trigger