Multi-Modal Track Reconstruction using Graph Neural Networks at Belle II
Lea Reuter 1, Tristan Brandes 1, Giacomo De Pietro 1,2 and Torben Ferber 1 · 2026
Read Multi-Modal Track Reconstruction using Graph Neural Networks at Belle II by Lea Reuter 1, Tristan Brandes 1, Giacomo De Pietro 1,2 and Torben Ferber 1 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-10897.txt.
Description
Lea Reuter and three co-authors extend multi-track reconstruction at Belle II to multiple detectors in a 2026 paper. High backgrounds and sensor ageing degrade track finding in the central drift chamber, reducing purity and speed. New recipes must mitigate that degradation as the test runs.
Building on an earlier end-to-end multi-track recipe, the team incorporates inputs from both the drift chamber and the silicon vertex sensor. Those inputs feed a multi-modal network. Graph neural networks handle the irregular sensor structure, while object condensation addresses the unknown, varying particle count per event. The approach at once identifies all tracks in an event and determines their settings, demonstrating improved reconstruction under Belle II settings.
Catalog record
| CREATOR | Lea Reuter 1, Tristan Brandes 1, Giacomo De Pietro 1,2 and Torben Ferber 1 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 16 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-10897