CaloChallenge 2022: a community challenge for fast calorimeter simulation

We present the results of the 'Fast Calorimeter Simulation Challenge 2022'-the CaloChallenge. We study state-of-the-art generative models on four calorimeter shower datasets of increasing dimensionality, ranging from a few hundred voxels to a few tens of thousand voxels. The 31 individual submissions span a wide range of current popular generative architectures, including variational autoencoders (VAEs), generative adversarial networks (GANs), normalizing flows, diffusion models, and models based on conditional flow matching. We compare all submissions in terms of quality of generated calorimeter showers, as well as shower generation time and model size. To assess the quality we use a broad range of different metrics including differences in one-dimensional histograms of observables, KPD/FPD scores, AUCs of binary classifiers, and the log-posterior of a multiclass classifier. The results of the CaloChallenge provide the most complete and comprehensive survey of cutting-edge approaches to calorimeter fast simulation to date. In addition, our work provides a uniquely detailed perspective on the important problem of how to evaluate generative models. As such, the results presented here should be applicable for other domains that use generative AI and require fast and faithful generation of samples in a large phase space.<b>Report Numbers</b>: HEPHY-ML-24-05, FERMILAB-PUB-24-0728-CMS, TTK-24-43.

Authors

Institutions

Publication Details

Journal
Reports on Progress in Physics
Published
2025-10-14
DOI
https://doi.org/10.1088/1361-6633/ae1304
Citations
6
Primary Topic
Machine Learning in Materials Science
Type
article
Field-Weighted Citation Impact
3.11
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

CaloChallenge 2022: a community challenge for fast calorimeter simulation

S. Vallecorsa, Claudius Krause, Mudhakar Srivatsa, Anthony L. Caterini et al.
6 citations
Reports on Progress in Physics
Machine Learning in Materials Science
3.11
article

CaloChallenge 2022: a community challenge for fast calorimeter simulation

S. Vallecorsa, Claudius Krause, Mudhakar Srivatsa, Anthony L. Caterini, Erik Buhmann, Jesse C. Cresswell, Matthew R. Buckley, T.Y. Kim, Renato Cardoso, J. A. Raine, Gabriel Loaiza-Ganem, B. Kaech, Tilman Plehn, Nathalie Soybelman, Peter P. McKeown, Luigi Favaro, Humberto Reyes-González, R. Kansal, N. Chernyavskaya, Kyongmin Yeo, M. Scham, Kalliopi Tsolaki, D. Salamani, Witold Pokorski, Eli Shlizerman, X.T. Liu, Vijay Ekambaram, Benjamin Nachman, V. M. Mikuni, Jayant Kalagnanam, Kristina Jaruskova, O. Amram, Ian Pang, E. Dreyer, Florian Ernst, David Shih, M. Franchini, Engin Eren, Sascha Diefenbacher, Marco Letizia, Anatolii Korol, M. Faucci Giannelli, Frank Gaede, Ayodele Ore, W. Korcari, E. Gross, Brendan Leigh Ross, Thorsten Buss, D. Kobylianskii, Federico Andrea Corchia, Shu Li, K. Krüger, Nam Nguyen, K. Borras, I.-A. Melzer-Pellmann, H. Qu, Kevin Pedro, T. Madula, S.‐C. Hsu, Q. Liu, L. Rinaldi, D. Krücker, Anna Zaborowska, Simon Schnake, G. Kasieczka, R. Zhang, Sofia Palacios Schweitzer, C. O. Shimmin, Piyush Raikwar
article en
6 citations

Abstract

We present the results of the 'Fast Calorimeter Simulation Challenge 2022'-the CaloChallenge. We study state-of-the-art generative models on four calorimeter shower datasets of increasing dimensionality, ranging from a few hundred voxels to a few tens of thousand voxels. The 31 individual submissions span a wide range of current popular generative architectures, including variational autoencoders (VAEs), generative adversarial networks (GANs), normalizing flows, diffusion models, and models based on conditional flow matching. We compare all submissions in terms of quality of generated calorimeter showers, as well as shower generation time and model size. To assess the quality we use a broad range of different metrics including differences in one-dimensional histograms of observables, KPD/FPD scores, AUCs of binary classifiers, and the log-posterior of a multiclass classifier. The results of the CaloChallenge provide the most complete and comprehensive survey of cutting-edge approaches to calorimeter fast simulation to date. In addition, our work provides a uniquely detailed perspective on the important problem of how to evaluate generative models. As such, the results presented here should be applicable for other domains that use generative AI and require fast and faithful generation of samples in a large phase space.<b>Report Numbers</b>: HEPHY-ML-24-05, FERMILAB-PUB-24-0728-CMS, TTK-24-43.

Reports on Progress in PhysicsVol. 88(11)
Rutgers, The State University of New Jersey (US), University of Geneva (CH), California Institute of Technology (US), Fermi National Accelerator Laboratory (US), IBM (United States) (US), Berkeley College (US), University of Wisconsin–Madison (US), Austrian Academy of Sciences (AT), Lawrence Berkeley National Laboratory (US), Universität Hamburg (DE), Forschungszentrum Jülich (DE), Shanghai Jiao Tong University (CN), University of Washington (US), Heidelberg University (DE), Deutsches Elektronen-Synchrotron DESY (DE), Yale University (US), University of California San Diego (US), Hochschule Campus Wien (AT), Kavli Institute for Particle Astrophysics and Cosmology (US), IBM Research - India (IN), Alliance for Safe Kids (US), Environmental and Occupational Health Sciences Institute (US), Istituto Nazionale di Fisica Nucleare, Roma Tor Vergata (IT), IBM Research - Thomas J. Watson Research Center (US), Istituto Nazionale di Fisica Nucleare, Sezione di Genova (IT), Institute of High Energy Physics (AT), Hamburg Institut (Germany) (DE), University of Washington Applied Physics Laboratory (US), Istituto Nazionale di Fisica Nucleare, Sezione di Bologna (IT), National Energy Research Scientific Computing Center (US), Institute of Natural Science (KP), University College London (GB), Toronto Metropolitan University (CA), Weizmann Institute of Science (IL), Chalmers University of Technology (SE), European Organization for Nuclear Research (CH), Heidelberg Institute for Theoretical Studies (DE), University of Genoa (IT), Nanjing University (CN), RWTH Aachen University (DE), University of Bologna (IT), University of California, Berkeley (US)
Openalex Percentile: Top 8%
Machine Learning in Materials Science
3.11
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.