19–23 May 2025
CERN
Europe/Zurich timezone

Track parameter regression with Transformers

Not scheduled
20m
61/1-201 - Pas perdus - Not a meeting room - (CERN)

61/1-201 - Pas perdus - Not a meeting room -

CERN

10
Show room on map
Poster 1 ML for object identification and reconstruction Poster Session

Speaker

Jeremy Couthures (Laboratoire d'Annecy de physique des particules, CNRS / Univ. Savoie Mont Blanc (FR))

Description

As the High-Luminosity LHC (HL-LHC) era approaches, significant improvements in reconstruction software are required to keep pace with the increased data rates and detector complexity. A persistent challenge for high-throughput GPU-based event reconstruction is the estimation of track parameters, which is traditionally performed using iterative Kalman Filter-based algorithms. While GPU-based track finding is progressing rapidly, the fitting stage remains a bottleneck. The main slowdown is coming from data movement between CPU and GPU which reduce the benefits of acceleration.

This work investigates a deep learning-based alternative using Transformer architectures for the prediction of the track parameters. The approach shows promising results on the TrackML dataset.

Would you like to be considered for an oral presentation? Yes

Authors

Alexis Vallier (L2I Toulouse, CNRS/IN2P3, UT3) Corentin Allaire (IJCLab, Université Paris-Saclay, CNRS/IN2P3) David Rousseau (IJCLab-Orsay) Jeremy Couthures (Laboratoire d'Annecy de physique des particules, CNRS / Univ. Savoie Mont Blanc (FR)) Marco Delmastro (CNRS/IN2P3 LAPP)

Presentation materials