Controlling Physical Attributes in GAN-Accelerated Simulation of Electromagnetic Calorimeters

TL;DR

This paper presents a GAN-based method for simulating electromagnetic calorimeters that incorporates physical attribute control, improving the conditioning and accuracy of particle shower generation in high-energy physics simulations.

Contribution

Introduces an auxiliary training task for GANs to enable attribute-aware conditioning in electromagnetic calorimeter simulations, enhancing physical fidelity.

Findings

Effective attribute conditioning in GAN-generated particle showers

Improved simulation speed over traditional methods

Potential for more accurate physics modeling

Abstract

High-precision modeling of subatomic particle interactions is critical for many fields within the physical sciences, such as nuclear physics and high energy particle physics. Most simulation pipelines in the sciences are computationally intensive -- in a variety of scientific fields, Generative Adversarial Networks have been suggested as a solution to speed up the forward component of simulation, with promising results. An important component of any simulation system for the sciences is the ability to condition on any number of physically meaningful latent characteristics that can effect the forward generation procedure. We introduce an auxiliary task to the training of a Generative Adversarial Network on particle showers in a multi-layer electromagnetic calorimeter, which allows our model to learn an attribute-aware conditioning mechanism.