Speaker
Description
We investigate the application of the Exa.TrkX pipeline---a graph neural network (GNN)-based tracking workflow originally developed for the High-Luminosity LHC---to particle tracking in a silicon-based tracker designed for a muon collider environment.
We adapt the Exa.TrkX workflow to identify signal muon track in this extremely dense environment. Using simulated datasets incorporating a generic all-silicon tracker with timing capabilities, together with beam-induced background overlays, we evaluate tracking efficiency, fake rate, and computational performance.
Our results demonstrate that the Exa.TrkX pipeline retains strong performance in this challenging regime, achieving excellent tracking efficiency while maintaining low fake rates. These findings highlight the potential of GNN-based approaches as a viable and scalable solution for track reconstruction at future muon collider experiments.
| What category does your poster fit in? | Other |
|---|