Enhancing events in neutrino telescopes through deep-learning-driven superresolution
利用深度学习超分辨率技术增强中微子望远镜事件重建
Recent discoveries by neutrino telescopes, such as the IceCube Neutrino Observatory, relied extensively on machine learning (ML) tools to infer physical quantities from the raw photon hits detected. Neutrino telescope reconstruction algorithms are limited by the sparse sampling of photons by the optical modules due to the relatively large spacing (10--100 m) between them. In this Letter, we propose a novel technique that learns photon transport through the detector medium through the use of deep-learning-driven superresolution of data events. These ``improved'' events can then be reconstructed using traditional or ML techniques, resulting in improved resolution. Our strategy arranges additional ``virtual'' optical modules within an existing detector geometry and trains a convolutional neural network to predict the hits on these virtual optical modules. We show that this technique improves the angular reconstruction of muons in a generic ice-based neutrino telescope. Our results readily extend to water-based neutrino telescopes and other event morphologies.
展开 ▾首次将图像超分辨率概念引入中微子望远镜,利用变分自编码器压缩时间序列,结合 UNet 预测虚拟 DOM 的光子信息,使 muon 角分辨率平均提升约 0.8°,且前向推断仅需约 0.3 ms。
中微子望远镜越来越依赖机器学习进行物理量重建,例如 IceCube 采用的卷积神经网络级联重建方法 Abbasi+ 2021 以及将最大似然与深度学习相结合的事件重建技术 Huennefeld+ 2021。随着探测器规模扩大和传感器间距增大,每个中微子事件所能获取的光子信息更加稀疏。本研究创新性地提出超分辨率策略,通过在现有探测器几何中插入“虚拟”光学模块,训练网络预测其光子到达时间序列,从而弥补稀疏采样造成的损失。该方法不仅适用于冰基望远镜,也可推广至水基望远镜如 KM3NeT,并可适应不同事件形态。未来,该技术中基于变分自编码器的时间序列压缩方案有望进一步用于压缩数据、加速重建,并集成到基于机器学习的事件生成模型中,从而为更大间距的下一代探测器提供高效重建手段,推动跨实验的通用重建框架发展。
刊出 2025-02-05 · 收录 2026-10-10