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

arXiv: https://arxiv.org/abs/2002.03605 · HTML: https://mlsanity.com/paper/2002.03605/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Feb 2020.

## Built on by (most cited first, 24 of 24)

- [MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks](https://mlsanity.com/paper/2101.08578/) (2021)
- [Graph Neural Networks for Charged Particle Tracking on FPGAs](https://mlsanity.com/paper/2112.02048/) (2021)
- [Machine Learning for Particle Flow Reconstruction at CMS](https://mlsanity.com/paper/2203.00330/) (2022)
- [Explaining machine-learned particle-flow reconstruction](https://mlsanity.com/paper/2111.12840/) (2021)
- [Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges](https://mlsanity.com/paper/2203.12852/) (2022)
- [Snowmass 2021 Computational Frontier CompF03 Topical Group Report: Machine Learning](https://mlsanity.com/paper/2209.07559/) (2022)
- [Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders](https://mlsanity.com/paper/2503.00131/) (2025)
- [High Pileup Particle Tracking with Object Condensation](https://mlsanity.com/paper/2312.03823/) (2023)
- [End-to-end multi-particle reconstruction in high occupancy imaging calorimeters with graph neural networks](https://mlsanity.com/paper/2204.01681/) (2022)
- [Do graph neural networks learn traditional jet substructure?](https://mlsanity.com/paper/2211.09912/) (2022)
- [Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors](https://mlsanity.com/paper/2309.06782/) (2023)
- [GeSS: Benchmarking Geometric Deep Learning under Scientific Applications with Distribution Shifts](https://mlsanity.com/paper/2310.08677/) (2023)
- [TrackFormers: In Search of Transformer-Based Particle Tracking for the High-Luminosity LHC Era](https://mlsanity.com/paper/2407.07179/) (2024)
- [Physics Instrument Design with Reinforcement Learning](https://mlsanity.com/paper/2412.10237/) (2024)
- [ColliderML: The First Release of an OpenDataDetector High-Luminosity Physics Benchmark Dataset](https://mlsanity.com/paper/2512.15230/) (2025)
- [End-to-end event reconstruction for precision physics at future colliders](https://mlsanity.com/paper/2603.04084/) (2026)
- [Deep Neural Network as an alternative to Boosted Decision Trees for PID](https://mlsanity.com/paper/2104.14045/) (2021)
- [What Machine Learning Can Do for Focusing Aerogel Detectors](https://mlsanity.com/paper/2312.02652/) (2023)
- [Exploring End-to-end Differentiable Neural Charged Particle Tracking -- A Loss Landscape Perspective](https://mlsanity.com/paper/2407.13420/) (2024)
- [Vision Calorimeter for Anti-neutron Reconstruction: A Baseline](https://mlsanity.com/paper/2408.10599/) (2024)
- [Denoising Graph Super-Resolution towards Improved Collider Event Reconstruction](https://mlsanity.com/paper/2409.16052/) (2024)
- [Tagging fully hadronic exotic decays of the vectorlike B quark using a graph neural network](https://mlsanity.com/paper/2505.07769/) (2025)
- [Contrastive Metric Learning for Point Cloud Segmentation in Highly Granular Detectors](https://mlsanity.com/paper/2603.23356/) (2026)
- [Using Line Segment Tracking to enable ML-based Track Reconstruction in CMS](https://mlsanity.com/paper/2610.15963/) (2026)

## Abstract (arXiv)

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.