DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: The HEP.TrkX Project: deep neural networks for HL-LHC online and offline tracking

Journal Article · · EPJ Web of Conferences (Online)
 [1];  [2];  [1];  [3];  [3];  [3];  [1];  [1];  [3];  [2];  [3];  [2];  [2]
  1. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
  2. California Inst. of Technology (CalTech), Pasadena, CA (United States)
  3. Fermi National Accelerator Lab. (FNAL), Batavia, IL (United States)

Particle track reconstruction in dense environments such as the detectors of the High Luminosity Large Hadron Collider (HL-LHC) is a challenging pattern recognition problem. Traditional tracking algorithms such as the combinatorial Kalman Filter have been used with great success in LHC experiments for years. However, these state-of-the-art techniques are inherently sequential and scale poorly with the expected increases in detector occupancy in the HL-LHC conditions. The HEP.TrkX project is a pilot project with the aim to identify and develop cross-experiment solutions based on machine learning algorithms for track reconstruction. Machine learning algorithms bring a lot of potential to this problem thanks to their capability to model complex non-linear data dependencies, to learn effective representations of high-dimensional data through training, and to parallelize easily on high-throughput architectures such as GPUs. This contribution will describe our initial explorations into this relatively unexplored idea space. Furthermore, we will discuss the use of recurrent (LSTM) and convolutional neural networks to find and fit tracks in toy detector data.

Research Organization:
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); Fermi National Accelerator Laboratory (FNAL), Batavia, IL (United States)
Sponsoring Organization:
USDOE Office of Science (SC), High Energy Physics (HEP) (SC-25)
Grant/Contract Number:
AC02-07CH11359
OSTI ID:
1375725
Report Number(s):
FERMILAB-CONF--17-326-CD; 1616026
Journal Information:
EPJ Web of Conferences (Online), Journal Name: EPJ Web of Conferences (Online) Vol. 150; ISSN 2100-014X
Publisher:
EDP SciencesCopyright Statement
Country of Publication:
United States
Language:
English

References (7)

ATLAS pixel detector electronics and sensors journal July 2008
Long Short-Term Memory journal November 1997
Deep learning journal May 2015
A concurrent track evolution algorithm for pattern recognition in the HERA-B main tracking system journal August 1997
Tracking at LHC journal September 2007
LHC Machine journal August 2008
The High-Luminosity upgrade of the LHC: Physics and Technology Challenges for the Accelerator and the Experiments journal April 2016

Cited By (3)

The HEP.TrkX Project: Deep Learning for Particle Tracking journal September 2018
FPGA-Accelerated Machine Learning Inference as a Service for Particle Physics Computing journal October 2019
Proton Tracking Algorithm in a Pixel-Based Range Telescope for Proton Computed Tomography preprint January 2020