PC-Droid: Faster diffusion and improved quality for particle cloud generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Leigh, Matthew, Sengupta, Debajyoti, Raine, John Andrew, Quétant, Guillaume, Golling, Tobias
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910338251751424
author Leigh, Matthew
Sengupta, Debajyoti
Raine, John Andrew
Quétant, Guillaume
Golling, Tobias
author_facet Leigh, Matthew
Sengupta, Debajyoti
Raine, John Andrew
Quétant, Guillaume
Golling, Tobias
contents Building on the success of PC-JeDi we introduce PC-Droid, a substantially improved diffusion model for the generation of jet particle clouds. By leveraging a new diffusion formulation, studying more recent integration solvers, and training on all jet types simultaneously, we are able to achieve state-of-the-art performance for all types of jets across all evaluation metrics. We study the trade-off between generation speed and quality by comparing two attention based architectures, as well as the potential of consistency distillation to reduce the number of diffusion steps. Both the faster architecture and consistency models demonstrate performance surpassing many competing models, with generation time up to two orders of magnitude faster than PC-JeDi and three orders of magnitude faster than Delphes.
format Preprint
id arxiv_https___arxiv_org_abs_2307_06836
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle PC-Droid: Faster diffusion and improved quality for particle cloud generation
Leigh, Matthew
Sengupta, Debajyoti
Raine, John Andrew
Quétant, Guillaume
Golling, Tobias
High Energy Physics - Experiment
Machine Learning
High Energy Physics - Phenomenology
Building on the success of PC-JeDi we introduce PC-Droid, a substantially improved diffusion model for the generation of jet particle clouds. By leveraging a new diffusion formulation, studying more recent integration solvers, and training on all jet types simultaneously, we are able to achieve state-of-the-art performance for all types of jets across all evaluation metrics. We study the trade-off between generation speed and quality by comparing two attention based architectures, as well as the potential of consistency distillation to reduce the number of diffusion steps. Both the faster architecture and consistency models demonstrate performance surpassing many competing models, with generation time up to two orders of magnitude faster than PC-JeDi and three orders of magnitude faster than Delphes.
title PC-Droid: Faster diffusion and improved quality for particle cloud generation
topic High Energy Physics - Experiment
Machine Learning
High Energy Physics - Phenomenology
url https://arxiv.org/abs/2307.06836