DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction
🤖 AI Summary
This work addresses the lack of standardized, publicly available datasets for machine learning–based track reconstruction in drift chambers, which has hindered fair comparison and reproducibility of algorithms. To bridge this gap, the authors introduce DCTracks, an open dataset generated via Monte Carlo simulation that includes both single-track and double-track events. They also propose dedicated evaluation metrics tailored specifically for track reconstruction tasks. This study establishes the first unified benchmarking platform that enables direct comparison of traditional algorithms and graph neural networks (GNNs) within a consistent framework, significantly enhancing reproducibility and laying a solid foundation for future methodological advancements in the field.
Application Category
📝 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.
Problem
Research questions and friction points this paper is trying to address.
drift chamber
track reconstruction
machine learning
open dataset
Monte Carlo simulation
Innovation
Methods, ideas, or system contributions that make the work stand out.
drift chamber
track reconstruction
machine learning
Graph Neural Networks
open dataset
🔎 Similar Papers
No similar papers found.
