Unsupervised Particle Tracking with Neuromorphic Computing

2025-02-10

AI Summary

Real-time charged particle track reconstruction in high-energy physics experiments—particularly for the CMS Phase II detector—faces severe challenges due to overwhelming combinatorial and accidental background noise under high-luminosity LHC conditions. Method: We propose an unsupervised spiking neural network (SNN) leveraging spike-timing-dependent plasticity (STDP) to directly learn spatiotemporal patterns from time-encoded detector hit signals, eliminating reliance on labeled training data. The network operates on the geometric layout and timing resolution of the CMS Phase II tracker. Contribution/Results: This work presents the first STDP-based unsupervised track identification framework validated on CMS Phase II geometry. It achieves significantly improved reconstruction accuracy in high-noise regimes compared to conventional approaches. Results demonstrate the feasibility and advantages of brain-inspired neuromorphic computing for low-power, real-time trigger systems in high-energy physics. The method establishes a novel edge-intelligence paradigm for future LHC upgrades, enabling efficient on-detector processing with minimal latency and energy consumption.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Computer Vision: Motion & Tracking

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We study the application of a neural network architecture for identifying charged particle trajectories via unsupervised learning of delays and synaptic weights using a spike-time-dependent plasticity rule. In the considered model, the neurons receive time-encoded information on the position of particle hits in a tracking detector for a particle collider, modeled according to the geometry of the Compact Muon Solenoid Phase II detector. We show how a spiking neural network is capable of successfully identifying in a completely unsupervised way the signal left by charged particles in the presence of conspicuous noise from accidental or combinatorial hits. These results open the way to applications of neuromorphic computing to particle tracking, motivating further studies into its potential for real-time, low-power particle tracking in future high-energy physics experiments.

Problem

• Unsupervised tracking of charged particles
• Neuromorphic computing in particle detectors
• Real-time, low-power particle identification

Innovation

• Unsupervised learning with neural networks
• Spike-time-dependent plasticity rule
• Neuromorphic computing for particle tracking