Object condensation: one-stage grid-free multi-object reconstruction in physics detectors, graph and image data

Abstract

High-energy physics detectors, images, and point clouds share many similarities in terms of object detection. However, while detecting an unknown number of objects in an image is well established in computer vision, even machine learning assisted object reconstruction algorithms in particle physics almost exclusively predict properties on an object-by-object basis. Traditional approaches from computer vision either impose implicit constraints on the object size or density and are not well suited for sparse detector data or rely on objects being dense and solid. The object condensation method proposed here is independent of assumptions on object size, sorting or object density, and further generalises to non-image-like data structures, such as graphs and point clouds, which are more suitable to represent detector signals. The pixels or vertices themselves serve as representations of the entire object, and a combination of learnable local clustering in a latent space and confidence assignment allows one to collect condensates of the predicted object properties with a simple algorithm. As proof of concept, the object condensation method is applied to a simple object classification problem in images and used to reconstruct multiple particles from detector signals. The latter results are also compared to a classic particle flow approach.

References in corpus (19)

• SSD: Single Shot MultiBox Detector
• scikit-image: Image processing in Python
• Particle-flow reconstruction and global event description with the CMS detector
• Particle Flow Calorimetry and the PandoraPFA Algorithm
• ParticleNet: Jet Tagging via Particle Clouds
• Deep Learning and its Application to LHC Physics
• Jet reconstruction and performance using particle flow with the ATLAS Detector
• The Machine Learning Landscape of Top Taggers
• Learning representations of irregular particle-detector geometry with distance-weighted graph networks
• JEDI-net: a jet identification algorithm based on interaction networks
• Pileup Mitigation with Machine Learning (PUMML)
• The Pandora Software Development Kit for Pattern Recognition
• Graph Neural Networks for Particle Reconstruction in High Energy Physics detectors
• Experimental Tests of Particle Flow Calorimetry
• Performance of Particle Flow Calorimetry at CLIC
• Towards a Computer Vision Particle Flow
• Electromagnetic Showers Beyond Shower Shapes
• Software compensation in Particle Flow reconstruction
• Topology classification with deep learning to improve real-time event selection at the LHC

Cited by in corpus (22)

• MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks
• Graph Neural Networks for Charged Particle Tracking on FPGAs
• Reconstructing particles in jets using set transformer and hypergraph prediction networks
• On the Use of Neural Networks for Energy Reconstruction in High-granularity Calorimeters
• Secondary Vertex Finding in Jets with Neural Networks
• Machine Learning for Particle Flow Reconstruction at CMS
• GPU coprocessors as a service for deep learning inference in high energy physics
• Fast convolutional neural networks for identifying long-lived particles in a high-granularity calorimeter
• Optimising longitudinal and lateral calorimeter granularity for software compensation in hadronic showers using deep neural networks
• Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors
• GNN-based end-to-end reconstruction in the CMS Phase 2 High-Granularity Calorimeter
• Portable acceleration of CMS computing workflows with coprocessors as a service
• End-to-End Multi-Track Reconstruction using Graph Neural Networks at Belle II
• Progress in End-to-End Optimization of Detectors for Fundamental Physics with Differentiable Programming
• Machine-learning-based prediction of parameters of secondaries in hadronic showers using calorimetric observables
• Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders
• Transforming a rare event search into a not-so-rare event search in real-time with deep learning-based object detection
• Accelerating Graph-based Tracking Tasks with Symbolic Regression
• TrackFormers: In Search of Transformer-Based Particle Tracking for the High-Luminosity LHC Era
• Tagging fully hadronic exotic decays of the vectorlike $\mathbf{B}$ quark using a graph neural network
• End-to-end multi-particle reconstruction in high occupancy imaging calorimeters with graph neural networks
• Real-time graph neural networks on FPGAs for the Belle II electromagnetic calorimeter