Deep-learning-based decomposition of overlapping-sparse images: application at the vertex of neutrino interactions

Abstract

Image decomposition plays a crucial role in various computer vision tasks, enabling the analysis and manipulation of visual content at a fundamental level. Overlapping images, which occur when multiple objects or scenes partially occlude each other, pose unique challenges for decomposition algorithms. The task intensifies when working with sparse images, where the scarcity of meaningful information complicates the precise extraction of components. This paper presents a solution that leverages the power of deep learning to accurately extract individual objects within multi-dimensional overlapping-sparse images, with a direct application in high-energy physics with decomposition of overlaid elementary particles obtained from imaging detectors. In particular, the proposed approach tackles a highly complex yet unsolved problem: identifying and measuring independent particles at the vertex of neutrino interactions, where one expects to observe detector images with multiple indiscernible overlapping charged particles. By decomposing the image of the detector activity at the vertex through deep learning, it is possible to infer the kinematic parameters of the identified low-momentum particles - which otherwise would remain neglected - and enhance the reconstructed energy resolution of the neutrino event. We also present an additional step - that can be tuned directly on detector data - combining the above method with a fully-differentiable generative model to improve the image decomposition further and, consequently, the resolution of the measured parameters, achieving unprecedented results. This improvement is crucial for precisely measuring the parameters that govern neutrino flavour oscillations and searching for asymmetries between matter and antimatter.

17 pages, 6 figures

References in corpus (22)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Evidence for oscillation of atmospheric neutrinos
• Direct Evidence for Neutrino Flavor Transformation from Neutral-Current Interactions in the Sudbury Neutrino Observatory
• The T2K Experiment
• Theory of Neutrinos: A White Paper
• Design and Construction of the MicroBooNE Detector
• Constraint on the Matter-Antimatter Symmetry-Violating Phase in Neutrino Oscillations
• Large Batch Optimization for Deep Learning: Training BERT in 76 minutes
• Identification of nuclear effects in neutrino-carbon interactions at low three-momentum transfer
• The NEUT Neutrino Interaction Simulation
• Pileup Mitigation with Machine Learning (PUMML)
• Improved constraints on neutrino mixing from the T2K experiment with $3.13\times 10^{21}$ protons on target
• The SuperFGD Prototype Charged Particle Beam Tests
• Image-based model parameter optimization using Model-Assisted Generative Adversarial Networks
• Antineutrino Charged-Current reactions on Hydrocarbon with Low Momentum Transfer
• Review of Liquid Argon Detector Technologies in the Neutrino Sector
• SARM: Sparse Autoregressive Model for Scalable Generation of Sparse Images in Particle Physics
• Study of final-state interactions of protons in neutrino-nucleus scattering with INCL and NuWro cascade models
• Measurement of inclusive charged-current $ν_{\numu}$ scattering on hydrocarbon at {<Enu>} 6 GeV with low three-momentum transfer
• SuperFGD prototype time resolution studies
• Multi-scale Image Decomposition using a Local Statistical Edge Model
• Differentiable programming for particle physics simulations

Cited by in corpus (1)

• Contrastive Learning for Robust Representations of Neutrino Data