Neutrino direction reconstructed using image-based CNNs

Researchers have developed a new method for reconstructing the direction of neutrinos detected by the IceCube Neutrino Observatory. This technique, termed "neutrino fingerprints," converts sparse detector data into compact images suitable for convolutional neural networks. A ResNet18 model utilizing these fingerprints achieved a mean angular error of 1.10 radians, demonstrating a competitive and interpretable approach to neutrino direction reconstruction. AI

IMPACT Introduces a novel image-based encoding for astrophysical data, potentially improving the efficiency and interpretability of deep learning models in scientific research.

RANK_REASON The cluster contains an academic paper detailing a new methodology for scientific data analysis. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• IceCube Neutrino Observatory
• neutrino fingerprints
• ResNet18

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →