-
Deep Learning for In-Orbit Cloud Segmentation and Classification in Hyperspectral Satellite Data
Authors:
Daniel Kovac,
Jan Mucha,
Jon Alvarez Justo,
Jiri Mekyska,
Zoltan Galaz,
Krystof Novotny,
Radoslav Pitonak,
Jan Knezik,
Jonas Herec,
Tor Arne Johansen
Abstract:
This article explores the latest Convolutional Neural Networks (CNNs) for cloud detection aboard hyperspectral satellites. The performance of the latest 1D CNN (1D-Justo-LiuNet) and two recent 2D CNNs (nnU-net and 2D-Justo-UNet-Simple) for cloud segmentation and classification is assessed. Evaluation criteria include precision and computational efficiency for in-orbit deployment. Experiments utili…
▽ More
This article explores the latest Convolutional Neural Networks (CNNs) for cloud detection aboard hyperspectral satellites. The performance of the latest 1D CNN (1D-Justo-LiuNet) and two recent 2D CNNs (nnU-net and 2D-Justo-UNet-Simple) for cloud segmentation and classification is assessed. Evaluation criteria include precision and computational efficiency for in-orbit deployment. Experiments utilize NASA's EO-1 Hyperion data, with varying spectral channel numbers after Principal Component Analysis. Results indicate that 1D-Justo-LiuNet achieves the highest accuracy, outperforming 2D CNNs, while maintaining compactness with larger spectral channel sets, albeit with increased inference times. However, the performance of 1D CNN degrades with significant channel reduction. In this context, the 2D-Justo-UNet-Simple offers the best balance for in-orbit deployment, considering precision, memory, and time costs. While nnU-net is suitable for on-ground processing, deployment of lightweight 1D-Justo-LiuNet is recommended for high-precision applications. Alternatively, lightweight 2D-Justo-UNet-Simple is recommended for balanced costs between timing and precision in orbit.
△ Less
Submitted 13 March, 2024;
originally announced March 2024.
-
Hybrid Quantum Classical Graph Neural Networks for Particle Track Reconstruction
Authors:
Cenk Tüysüz,
Carla Rieger,
Kristiane Novotny,
Bilge Demirköz,
Daniel Dobos,
Karolos Potamianos,
Sofia Vallecorsa,
Jean-Roch Vlimant,
Richard Forster
Abstract:
The Large Hadron Collider (LHC) at the European Organisation for Nuclear Research (CERN) will be upgraded to further increase the instantaneous rate of particle collisions (luminosity) and become the High Luminosity LHC (HL-LHC). This increase in luminosity will significantly increase the number of particles interacting with the detector. The interaction of particles with a detector is referred to…
▽ More
The Large Hadron Collider (LHC) at the European Organisation for Nuclear Research (CERN) will be upgraded to further increase the instantaneous rate of particle collisions (luminosity) and become the High Luminosity LHC (HL-LHC). This increase in luminosity will significantly increase the number of particles interacting with the detector. The interaction of particles with a detector is referred to as "hit". The HL-LHC will yield many more detector hits, which will pose a combinatorial challenge by using reconstruction algorithms to determine particle trajectories from those hits. This work explores the possibility of converting a novel Graph Neural Network model, that can optimally take into account the sparse nature of the tracking detector data and their complex geometry, to a Hybrid Quantum-Classical Graph Neural Network that benefits from using Variational Quantum layers. We show that this hybrid model can perform similar to the classical approach. Also, we explore Parametrized Quantum Circuits (PQC) with different expressibility and entangling capacities, and compare their training performance in order to quantify the expected benefits. These results can be used to build a future road map to further develop circuit based Hybrid Quantum-Classical Graph Neural Networks.
△ Less
Submitted 26 September, 2021;
originally announced September 2021.
-
Performance of Particle Tracking Using a Quantum Graph Neural Network
Authors:
Cenk Tüysüz,
Kristiane Novotny,
Carla Rieger,
Federico Carminati,
Bilge Demirköz,
Daniel Dobos,
Fabio Fracas,
Karolos Potamianos,
Sofia Vallecorsa,
Jean-Roch Vlimant
Abstract:
The Large Hadron Collider (LHC) at the European Organisation for Nuclear Research (CERN) will be upgraded to further increase the instantaneous rate of particle collisions (luminosity) and become the High Luminosity LHC. This increase in luminosity, will yield many more detector hits (occupancy), and thus measurements will pose a challenge to track reconstruction algorithms being responsible to de…
▽ More
The Large Hadron Collider (LHC) at the European Organisation for Nuclear Research (CERN) will be upgraded to further increase the instantaneous rate of particle collisions (luminosity) and become the High Luminosity LHC. This increase in luminosity, will yield many more detector hits (occupancy), and thus measurements will pose a challenge to track reconstruction algorithms being responsible to determine particle trajectories from those hits. This work explores the possibility of converting a novel Graph Neural Network model, that proven itself for the track reconstruction task, to a Hybrid Graph Neural Network in order to benefit the exponentially growing Hilbert Space. Several Parametrized Quantum Circuits (PQC) are tested and their performance against the classical approach is compared. We show that the hybrid model can perform similar to the classical approach. We also present a future road map to further increase the performance of the current hybrid model.
△ Less
Submitted 25 January, 2021; v1 submitted 2 December, 2020;
originally announced December 2020.
-
A Quantum Graph Neural Network Approach to Particle Track Reconstruction
Authors:
Cenk Tüysüz,
Federico Carminati,
Bilge Demirköz,
Daniel Dobos,
Fabio Fracas,
Kristiane Novotny,
Karolos Potamianos,
Sofia Vallecorsa,
Jean-Roch Vlimant
Abstract:
Unprecedented increase of complexity and scale of data is expected in computation necessary for the tracking detectors of the High Luminosity Large Hadron Collider (HL-LHC) experiments. While currently used Kalman filter based algorithms are reaching their limits in terms of ambiguities from increasing number of simultaneous collisions, occupancy, and scalability (worse than quadratic), a variety…
▽ More
Unprecedented increase of complexity and scale of data is expected in computation necessary for the tracking detectors of the High Luminosity Large Hadron Collider (HL-LHC) experiments. While currently used Kalman filter based algorithms are reaching their limits in terms of ambiguities from increasing number of simultaneous collisions, occupancy, and scalability (worse than quadratic), a variety of machine learning approaches to particle track reconstruction are explored. It has been demonstrated previously by HEP.TrkX using TrackML datasets, that graph neural networks, by processing events as a graph connecting track measurements can provide a promising solution by reducing the combinatorial background to a manageable amount and are scaling to a computationally reasonable size. In previous work, we have shown a first attempt of Quantum Computing to Graph Neural Networks for track reconstruction of particles. We aim to leverage the capability of quantum computing to evaluate a very large number of states simultaneously and thus to effectively search a large parameter space. As the next step in this paper, we present an improved model with an iterative approach to overcome the low accuracy convergence of the initial oversimplified Tree Tensor Network (TTN) model.
△ Less
Submitted 14 July, 2020;
originally announced July 2020.
-
Particle Track Reconstruction with Quantum Algorithms
Authors:
Cenk Tüysüz,
Federico Carminati,
Bilge Demirköz,
Daniel Dobos,
Fabio Fracas,
Kristiane Novotny,
Karolos Potamianos,
Sofia Vallecorsa,
Jean-Roch Vlimant
Abstract:
Accurate determination of particle track reconstruction parameters will be a major challenge for the High Luminosity Large Hadron Collider (HL-LHC) experiments. The expected increase in the number of simultaneous collisions at the HL-LHC and the resulting high detector occupancy will make track reconstruction algorithms extremely demanding in terms of time and computing resources. The increase in…
▽ More
Accurate determination of particle track reconstruction parameters will be a major challenge for the High Luminosity Large Hadron Collider (HL-LHC) experiments. The expected increase in the number of simultaneous collisions at the HL-LHC and the resulting high detector occupancy will make track reconstruction algorithms extremely demanding in terms of time and computing resources. The increase in number of hits will increase the complexity of track reconstruction algorithms. In addition, the ambiguity in assigning hits to particle tracks will be increased due to the finite resolution of the detector and the physical closeness of the hits. Thus, the reconstruction of charged particle tracks will be a major challenge to the correct interpretation of the HL-LHC data. Most methods currently in use are based on Kalman filters which are shown to be robust and to provide good physics performance. However, they are expected to scale worse than quadratically. Designing an algorithm capable of reducing the combinatorial background at the hit level, would provide a much cleaner initial seed to the Kalman filter, strongly reducing the total processing time. One of the salient features of Quantum Computers is the ability to evaluate a very large number of states simultaneously, making them an ideal instrument for searches in a large parameter space. In fact, different R\&D initiatives are exploring how Quantum Tracking Algorithms could leverage such capabilities. In this paper, we present our work on the implementation of a quantum-based track finding algorithm aimed at reducing combinatorial background during the initial seeding stage. We use the publicly available dataset designed for the kaggle TrackML challenge.
△ Less
Submitted 18 March, 2020;
originally announced March 2020.
-
Higgs Decay into Two Photons at the Boundary of a Warped Extra Dimension
Authors:
Juliane Hahn,
Clara Hörner,
Raoul Malm,
Matthias Neubert,
Kristiane Novotny,
Christoph Schmell
Abstract:
A detailed five-dimensional calculation of the Higgs-boson decay into two photons is performed in both the minimal and the custodially protected Randall-Sundrum (RS) model, where the Standard Model (SM) fields propagate in the bulk and the scalar sector lives on or near the IR brane. It is explicitly shown that the $R_ξ$ gauge invariance of the sum of diagrams involving bosonic fields in the SM al…
▽ More
A detailed five-dimensional calculation of the Higgs-boson decay into two photons is performed in both the minimal and the custodially protected Randall-Sundrum (RS) model, where the Standard Model (SM) fields propagate in the bulk and the scalar sector lives on or near the IR brane. It is explicitly shown that the $R_ξ$ gauge invariance of the sum of diagrams involving bosonic fields in the SM also applies to the case of these RS scenarios. An exact expression for the $h\toγγ$ amplitude in terms of the five-dimensional (5D) gauge-boson and fermion propagators is presented, which includes the full dependence on the Higgs-boson mass. Closed expressions for the 5D $W$-boson propagators in the minimal and the custodial RS model are derived, which are valid to all orders in $v^2/M_\textrm{KK}^2$. In contrast to the fermion case, the result for the bosonic contributions to the $h\toγγ$ amplitude is insensitive to the details of the localization of the Higgs profile on or near the IR brane. The various RS predictions for the rate of the $pp\to h\toγγ$ process are compared with the latest LHC data, and exclusion regions for the RS model parameters are derived.
△ Less
Submitted 7 February, 2014; v1 submitted 19 December, 2013;
originally announced December 2013.
-
5D Perspective on Higgs Production at the Boundary of a Warped Extra Dimension
Authors:
Raoul Malm,
Matthias Neubert,
Kristiane Novotny,
Christoph Schmell
Abstract:
A comprehensive, five-dimensional calculation of Higgs-boson production in gluon fusion is performed for both the minimal and the custodially protected Randall-Sundrum (RS) model, with Standard Model fields propagating in the bulk and the scalar sector confined on or near the IR brane. For the first time, an exact expression for the gg->h amplitude in terms of the five-dimensional fermion propagat…
▽ More
A comprehensive, five-dimensional calculation of Higgs-boson production in gluon fusion is performed for both the minimal and the custodially protected Randall-Sundrum (RS) model, with Standard Model fields propagating in the bulk and the scalar sector confined on or near the IR brane. For the first time, an exact expression for the gg->h amplitude in terms of the five-dimensional fermion propagator is derived, which includes the full dependence on the Higgs-boson mass. Various results in the literature are reconciled and shown to correspond to different incarnations of the RS model, in which the Higgs field is either localized on the IR brane or is described in terms of a narrow bulk state. The results in the two scenarios differ in a qualitative way: the gg->h amplitude is suppressed in models where the scalar sector is localized on the IR brane, while it tends to be enhanced in bulk Higgs models. In both cases, effects of higher-dimensional operators contributing to the gg->h amplitude at tree level are shown to be numerically suppressed under reasonable assumptions. There is no smooth cross-over between the two scenarios, since the effective field-theory description breaks down in the transition region. A detailed phenomenological analysis of Higgs production in various RS scenarios is presented, and for each scenario the regions of parameter space already excluded by LHC data are derived.
△ Less
Submitted 24 December, 2013; v1 submitted 22 March, 2013;
originally announced March 2013.
