Search papers, labs, and topics across Lattice.
The paper introduces DCTracks, a new Monte Carlo dataset of drift chamber events designed to facilitate the development and evaluation of machine learning-based track reconstruction algorithms. The dataset includes both single- and two-track events, enabling researchers to train and test algorithms on realistic simulated data. The authors establish standardized metrics for track reconstruction performance and provide baseline results using traditional algorithms and a GNN method, promoting reproducible research.
A new open dataset with standardized metrics finally lets researchers rigorously compare machine learning approaches for drift chamber track reconstruction.
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.
