Particle track reconstruction with the TrackNETv2

Goncharov, Pavel; Ososkov, G.; Baranov, Dmitriy

doi:10.1063/1.5130102

Cited by 8 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Deep learning reconstruction methods: With the availability of modern deep learning architectures together with dedicated computing hardware, deep learning demonstrated impressive performance in particle tracking due to the ability to learn from raw data with no or minimal assumptions about the underlying system. Early approaches heavily utilized LSTM [29], [30], [31] and CNN [30], [31] architectures, and were later on superseded by graph neural networks [1], [2], sparsely capturing relations between particle hits. Here, most approaches rely on an edge classification scheme together with a final planning module extracting feasible track candidates leveraging predicted edge scores [29], [32], [33], [34].…”

Section: Related Work and Motivationmentioning

confidence: 99%

Towards Neural Charged Particle Tracking in Digital Tracking Calorimeters With Reinforcement Learning

Kortus

Gauger

2023

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

We propose a novel technique for reconstructing charged particles in digital tracking calorimeters using reinforcement learning aiming to benefit from the rapid progress and success of neural network architectures without the dependency on simulated or manually-labeled data. Here we optimize by trial-and-error a behavior policy acting as an approximation to the full combinatorial optimization problem, maximizing the physical plausibility of sampled trajectories. In modern processing pipelines used in high energy physics and related applications, tracking plays an essential role allowing to identify and follow charged particle trajectories traversing particle detectors. Due to the high multiplicity of charged particles and their physical interactions, randomly deflecting the particles, the reconstruction is a challenging undertaking, requiring fast, accurate and robust algorithms. Our approach works on graph-structured data, capturing track hypotheses through edge connections between particles in the detector layers. We demonstrate in a comprehensive study on simulated data for a particle detector used for proton computed tomography, the high potential as well as the competitiveness of our approach compared to a heuristic search algorithm and a model trained on ground truth. Finally, we point out limitations of our approach, guiding towards a robust foundation for further development of reinforcement learning based tracking.

show abstract

Section: Related Work and Motivationmentioning

confidence: 99%

Towards Neural Charged Particle Tracking in Digital Tracking Calorimeters With Reinforcement Learning

Kortus

Gauger

2023

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…Having performed a combinatorial spatial search for track candidates simultaneously using two coordinate projections, one can use then a recurrent neural network (RNN) to separate true and false (ghost) tracks. This idea underlies the TrackNET group of algorithms, including TrackNETv1 [10], TrackNETv2 [11], and TrackNETv3, where each following algorithm is an evolution of the previous one. Since the first two algorithms have been described elsewhere, we only briefly summarize their properties below.…”

Section: Tracknetv3 Neural Networkmentioning

confidence: 99%

Deep learning for track recognition in pixel and strip-based particle detectors

et al. 2022

View full text Add to dashboard Cite

The reconstruction of charged particle trajectories in tracking detectors is a key problem in the analysis of experimental data for high-energy and nuclear physics. The amount of data in modern experiments is so large that classical tracking methods, such as the Kalman filter, cannot process them fast enough. To solve this problem, we developed two neural network algorithms based on deep learning architectures for track recognition in pixel and strip-based particle detectors. These are TrackNETv3 for local (track by track) and RDGraphNet for global (all tracks in an event) tracking. These algorithms were tested using the GEM tracker of the BM@N experiment at JINR (Dubna) and the cylindrical GEM inner tracker of the BESIII experiment at IHEP CAS (Beijing). The RDGraphNet model, based on a reverse directed graph, showed encouraging results: 95% recall and 74% precision for track finding. The TrackNETv3 model demonstrated a recall value of 95% and 76% precision. This result can be improved after further model optimization.

show abstract

“…В процессе совершенствования модели TrackNET авторы данной статьи пришли к мысли о том, что процедура предсказания эллиптической области для поиска следующего хита трека уже содержит необходимую информацию о гладкости кривой. Эти выводы привели к созданию второй версии модели TrackNETv2 [Goncharov et al, 2019], описание которой приведено ниже в данной статье. TrackNETv2 является наиболее перспективным алгоритмом локального адаптивного трекинга, так как по сути своего функционирования эта модель выполняет роль обучаемого нейросетевого фильтра Калмана.…”

Section: предыдущие исследованияunclassified

“…Методы TrackNETv2 и GNN имеют большой потенциал для реконструкции треков в эксперименте BESIII в условиях столь сильного зашумления, а также показывают перспективные результаты при тестировании на данных Монте-Карло-моделирования для эксперимента BM@N с аналогичным типом детектора [Goncharov et al, 2019;Shchavelev et al, 2019;Goncharov et al, 2020].…”

Section: внутренний трекер эксперимента Besiiiunclassified

“…Нейросетевая модель для трекинга, TrackNETv2, изначально была предложена для реконструкции треков в эксперименте BM@N с фиксированной мишенью [Goncharov et al, 2019].…”

Section: особенности Tracknetv2 для цилиндрических трековых детекторовunclassified

See 1 more Smart Citation