DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction

TL;DR

This paper presents a new open dataset for drift chamber track reconstruction, enabling standardized evaluation of traditional and ML-based algorithms, including GNNs, to improve future research in particle tracking.

Contribution

It introduces a Monte Carlo dataset and evaluation metrics for ML-based drift chamber track reconstruction, including a GNN approach, promoting reproducibility and comparison.

Findings

GNNs outperform traditional algorithms in certain scenarios

Standardized metrics enable fair comparison of methods

Open dataset facilitates future research and benchmarking

Abstract

We introduce a Monte Carlo (MC) dataset of single- and two-track drift chamber events to advance Machine Learning (ML)-based track reconstruction. To enable standardized and comparable evaluation, we define track reconstruction specific metrics and report results for traditional track reconstruction algorithms and a Graph Neural Networks (GNNs) method, facilitating rigorous, reproducible validation for future research.