Enhancing Photon Identification with Neural Network Methods

TL;DR

This paper demonstrates that residual convolutional neural networks, especially ResNet architectures with physics-informed loss functions, significantly improve photon identification in high-luminosity collider environments with overlapping electromagnetic showers.

Contribution

The study introduces a ResNet-based CNN approach for photon identification that outperforms traditional BDTs and DNNs, especially in complex overlapping shower scenarios.

Findings

ResNet-based CNNs outperform BDTs and DNNs in photon identification.

Augmenting ResNet with soft scoring and $ riangle R$ regression further improves accuracy.

Residual architectures with physics-informed loss functions are effective in high-luminosity collider environments.

Abstract

We investigate photon--pion discrimination in regimes where electromagnetic showers overlap at the scale of calorimeter granularity. Using full detector simulations with fine-grained calorimeter segmentation of approximately $0.025\times0.025$ in $(\eta,\phi)$, we benchmark three approaches: boosted decision trees (BDTs) on shower-shape variables, dense neural networks (DNNs) on the same features, and a ResNet-based convolutional neural network operating directly on calorimeter cell energies. The ResNet significantly outperformed both baseline methods, achieving further gains when augmented with soft scoring and an auxiliary $\Delta R$ regression head. Our results demonstrate that residual convolutional architectures, combined with physics-informed loss functions, can substantially improve photon identification in high-luminosity collider environments in which overlapping electromagnetic showers challenge traditional methods.