arXiv subjects

Thomas Lang

Publications and source records attributed to Thomas Lang.

At least 19 recordsLinked to original sources

arXiv2026

QCxSimulation: Scatter-Aware X-Ray Projection Radiography via Discrete-Time Quantum Walks

X-ray projection radiography is a non-invasive imaging technique used in medical diagnostics and industrial inspection. The simulation of X-ray projections is commonly used to optimise acquisition protocols and improve image quality before performing costly scans. Classical photon transport simulations that include realistic X-ray scattering physics are computationally expensive because they require the sampling of a large number of distinct scattering paths. This limits the practical exploration of parameter spaces such as beam energy. Quantum computing offers the potential to solve high-dimensional problems faster by making use of quantum properties such as superposition. This work introduces a discrete-time quantum walk algorithm that simulates the transport of X-ray photons through heterogeneous volumes. It approximates the physics of X-ray projection radiography, including processes such as photoelectric absorption and higher-order scattering, including Compton and Rayleigh scattering. The quantum walk encodes all admissible photon paths into a single quantum state, enabling all scattering histories to be propagated simultaneously via the superposition principle. This quantum state representation enables flexible readout of various imaging modalities, including the primary, i.e., unscattered, image, or images exclusively containing Rayleigh and Compton scattering of specified orders. A quantitative comparison with classically computed reference simulations shows that the proposed quantum walk accurately reproduces radiographic projections, given the limitations of the underlying physical model. These results indicate that quantum circuits for X-ray transport can produce accurate radiographic images and imply that, as quantum hardware scales up, these algorithms could outperform classical Monte Carlo-based approaches in large-scale, scatter-aware virtual imaging studies.

quant-ph↗

arXiv2025

Understanding and Improving Hyperbolic Deep Reinforcement Learning

The exponential volume growth of hyperbolic geometry can embed the hierarchical relationships between states in reinforcement learning (RL) with far less distortion than Euclidean space. However, hyperbolic deep RL faces severe optimization challenges, and formal analysis of why optimization fails is lacking. We identify key factors that determine the success and failure of training hyperbolic deep RL agents. By analyzing the gradients of core operations in the Poincar\'e Ball and Hyperboloid models of hyperbolic geometry, we show that large-norm embeddings destabilize gradient-based training, leading to trust-region violations in proximal policy optimization (PPO). Based on these insights, we introduce Hyper++, a new hyperbolic deep RL agent that consists of three components: (1) feature regularization guaranteeing bounded norms while avoiding the curse of dimensionality from clipping; (2) a categorical value loss for stable critic training; and (3) a more optimization-friendly formulation of hyperbolic network layers. On ProcGen, we show that Hyper++ guarantees stable learning, outperforms prior hyperbolic agents, and reduces wall-clock time by approximately 30%. On Atari-5 with Double DQN, Hyper++ strongly outperforms Euclidean and hyperbolic baselines. We release our code at https://github.com/Probabilistic-and-Interactive-ML/hyper-rl.

cs.LG↗

arXiv2025

Exploring Large Quantities of Secondary Data from High-Resolution Synchrotron X-ray Computed Tomography Scans Using AccuStripes

The analysis of secondary quantitative data extracted from high-resolution synchrotron X-ray computed tomography scans represents a significant challenge for users. While a number of methods have been introduced for processing large three-dimensional images in order to generate secondary data, there are only a few techniques available for simple and intuitive visualization of such data in their entirety. This work employs the AccuStripes visualization technique for that purpose, which enables the visual analysis of secondary data represented by an ensemble of univariate distributions. It supports different schemes for adaptive histogram binnings in combination with several ways of rendering aggregated data and it allows the interactive selection of optimal visual representations depending on the data and the use case. We demonstrate the usability of AccuStripes on a high-resolution synchrotron scan of a particle-reinforced metal matrix composite sample, containing more than 20 million particles. Through AccuStripes, detailed insights are facilitated into distributions of derived particle characteristics of the entire sample. Furthermore, research questions such as how the overall shape of the particles is or how homogeneously they are distributed across the sample can be answered.

cs.HC↗

arXiv2025

Quantum Image Visualizer: Visual Debugging of Quantum Image Processing Circuits

Quantum computing is an emerging field that utilizes the unique principles of quantum mechanics to offer significant advantages in algorithm execution over classical approaches. This potential is particularly promising in the domain of quantum image processing, which aims to manipulate all pixels simultaneously. However, the process of designing and verifying these algorithms remains a complex and error-prone task. To address this challenge, new methods are needed to support effective debugging of quantum circuits. The Quantum Image Visualizer is an interactive visual analysis tool that allows for the examination of quantum images and their transformation throughout quantum circuits. The framework incorporates two overview visualizations that trace image evolution across a sequence of gates based on the most probable outcomes. Interactive exploration allows users to focus on relevant gates, and select pixels of interest. Upon selection, detailed visualizations enable in-depth inspection of individual pixels and their probability distributions, revealing how specific gates influence the likelihood of pixel color values and the magnitude of these changes. An evaluation of the Quantum Image Visualizer was conducted through in-depth interviews with eight domain experts. The findings demonstrate the effectiveness and practical value of our approach in supporting visual debugging of quantum image processing circuits.

cs.HC↗

arXiv2024

Breaking the Reclustering Barrier in Centroid-based Deep Clustering

This work investigates an important phenomenon in centroid-based deep clustering (DC) algorithms: Performance quickly saturates after a period of rapid early gains. Practitioners commonly address early saturation with periodic reclustering, which we demonstrate to be insufficient to address performance plateaus. We call this phenomenon the "reclustering barrier" and empirically show when the reclustering barrier occurs, what its underlying mechanisms are, and how it is possible to Break the Reclustering Barrier with our algorithm BRB. BRB avoids early over-commitment to initial clusterings and enables continuous adaptation to reinitialized clustering targets while remaining conceptually simple. Applying our algorithm to widely-used centroid-based DC algorithms, we show that (1) BRB consistently improves performance across a wide range of clustering benchmarks, (2) BRB enables training from scratch, and (3) BRB performs competitively against state-of-the-art DC algorithms when combined with a contrastive loss. We release our code and pre-trained models at https://github.com/Probabilistic-and-Interactive-ML/breaking-the-reclustering-barrier .

cs.LG↗

arXiv2024

Representation of Classical Data on Quantum Computers

Quantum computing is currently gaining significant attention, not only from the academic community but also from industry, due to its potential applications across several fields for addressing complex problems. For any practical problem which may be tackled using quantum computing, it is imperative to represent the data used onto a quantum computing system. Depending on the application, many different types of data and data structures occur, including regular numbers, higher-dimensional data structures, e.g., n-dimensional images, up to graphs. This report aims to provide an overview of existing methods for representing these data types on gate-based quantum computers.

quant-ph↗

arXiv2023

The Influence of Satellite Trails on H.E.S.S. Gamma-Ray Astronomical Observations

The number of satellites launched into low earth orbit has almost tripled (to over 4000) in the last three years due to the increasing commercialisation of space. Satellite constellations with a total of over 400,000 satellites are proposed to be launched in the near future. Many of these satellites are highly reflective, resulting in a high optical brightness that affects ground-based astronomical observations across the electromagnetic spectrum. Despite this, the potential effect of these satellites on Imaging Atmospheric Cherenkov Telescopes (IACTs) has so far been assumed to be negligible due to their nanosecond integration times. This has, however, never been verified. We aim to identify satellite trails in data taken by the High Energy Stereoscopic System (H.E.S.S.) IACT array in Namibia, using Night Sky Background (NSB) data from the CT5 camera installed in 2019. We determine which observation times and pointing directions are affected the most, and evaluate the impact on Hillas parameters used for classification and reconstruction of high-energy Extensive Air Shower events. Finally, we predict how future planned satellite launches will affect gamma-ray observations with IACTs.

astro-ph.IM↗

arXiv2023

Impact of Satellite Trails on H.E.S.S. Astronomical Observations

The number of satellites launched into Earth's orbit has almost tripled in the last three years due to the increasing commercialisation of space. Multiple satellite constellations, consisting of over 400,000 individual satellites, have either been partially launched or are proposed for launch in the near future. Many of these satellites are highly reflective, resulting in a high optical brightness that affects ground-based astronomical observations. Despite this caveat, the potential effect of these satellites on gamma-ray-observing Imaging Atmospheric Cherenkov Telescopes (IACTs) has largely been assumed to be negligible due to their nanosecond-scale integration times. However, this assumption has not been verified to date. As IACTs are sensitive to optical wavelength light, we aim to identify satellite trails in data taken by the High Energy Stereoscopic System (H.E.S.S.) IACT array. In particular, this study is aimed at quantifying the potential effects on data quality and extensive air shower event classification and reconstruction. Using night sky background measurements from H.E.S.S., we determined which observation times and pointing directions are affected most by these satellite trails. We then evaluated their impact on the standard Hillas parameter variables used for event analysis. Due to the brightest trails, false trigger events can occur, however, for most modern analyses, the effect on astronomical results will be minimal. We observe a mild increase in the rate of trail detections over time, which is partially correlated with the number of satellite launches. Overall, the fraction of H.E.S.S. data affected is currently minimal. We note that these trails could still have a non-negligible effect on future Cherenkov Telescope Array observations if advanced analysis techniques designed to lower the energy threshold of the instrument are applied.

astro-ph.IM↗

arXiv2023

Clustering large 3D volumes: A sampling-based approach

In many applications of X-ray computed tomography, an unsupervised segmentation of the reconstructed 3D volumes forms an important step in the image processing chain for further investigation of the digitized object. Therefore, the goal is to train a clustering algorithm on the volume, which produces a voxelwise classification by assigning a cluster index to each voxel. However, clustering methods, e.g., K-Means, typically have an asymptotic polynomial runtime with respect to the dataset size, and thus, these techniques are rarely applicable to large volumes. In this work, we introduce a novel clustering technique based on random sampling, which allows for the voxelwise classification of arbitrarily large volumes. The presented method conducts efficient linear passes over the data to extract a representative random sample of a fixed size on which the classifier can be trained. Then, a final linear pass performs the segmentation and assigns a cluster index to each individual voxel. Quantitative and qualitative evaluations show that excellent results can be achieved even with a very small sample size. Consequently, the unsupervised segmentation by means of clustering becomes feasible for arbitrarily large volumes.

cs.CV↗

arXiv2022

Geometric Active Learning for Segmentation of Large 3D Volumes

Segmentation, i.e., the partitioning of volumetric data into components, is a crucial task in many image processing applications ever since such data could be generated. Most existing applications nowadays, specifically CNNs, make use of voxelwise classification systems which need to be trained on a large number of annotated training volumes. However, in many practical applications such data sets are seldom available and the generation of annotations is time-consuming and cumbersome. In this paper, we introduce a novel voxelwise segmentation method based on active learning on geometric features. Our method uses interactively provided seed points to train a voxelwise classifier based entirely on local information. The combination of an ad hoc incorporation of domain knowledge and local processing results in a flexible yet efficient segmentation method that is applicable to three-dimensional volumes without size restrictions. We illustrate the potential and flexibility of our approach by applying it to selected computed tomography scans where we perform different segmentation tasks to scans from different domains and of different sizes.

cs.CV↗

arXiv2022

Feature-Adaptive Interactive Thresholding of Large 3D Volumes

Thresholding is the most widely used segmentation method in volumetric image processing, and its pointwise nature makes it attractive for the fast handling of large three-dimensional samples. However, global thresholds often do not properly extract components in the presence of artifacts, measurement noise or grayscale value fluctuations. This paper introduces Feature-Adaptive Interactive Thresholding (FAITH), a thresholding technique that incorporates (geometric) features, local processing and interactive user input to overcome these limitations. Given a global threshold suitable for most regions, FAITH uses interactively selected seed voxels to identify critical regions in which that threshold will be adapted locally on the basis of features computed from local environments around these voxels. The combination of domain expert knowledge and a rigorous mathematical model thus enables a very exible way of local thresholding with intuitive user interaction. A qualitative analysis shows that the proposed model is able to overcome limitations typically occuring in plain thresholding while staying efficient enough to also allow the segmentation of big volumes.

cs.CV↗

arXiv2022

A New Hip Fracture Risk Index Derived from FEA-Computed Proximal Femur Fracture Loads and Energies-to-Failure

Hip fracture risk assessment is an important but challenging task. Quantitative CT-based patient specific finite element analysis (FEA) computes the force (fracture load) to break the proximal femur in a particular loading condition. It provides different structural information about the proximal femur that can influence a subject overall fracture risk. To obtain a more robust measure of fracture risk, we used principal component analysis (PCA) to develop a global FEA computed fracture risk index that incorporates the FEA-computed yield and ultimate failure loads and energies to failure in four loading conditions (single-limb stance and impact from a fall onto the posterior, posterolateral, and lateral aspects of the greater trochanter) of 110 hip fracture subjects and 235 age and sex matched control subjects from the AGES-Reykjavik study. We found that the first PC (PC1) of the FE parameters was the only significant predictor of hip fracture. Using a logistic regression model, we determined if prediction performance for hip fracture using PC1 differed from that using FE parameters combined by stratified random resampling with respect to hip fracture status. The results showed that the average of the area under the receive operating characteristic curve (AUC) using PC1 was always higher than that using all FE parameters combined in the male subjects. The AUC of PC1 and AUC of the FE parameters combined were not significantly different than that in the female subjects or in all subjects

cs.LG↗

arXiv2013

Correlated D-meson decays competing against thermal QGP dilepton radiation

The QGP that might be created in ultrarelativistic heavy-ion collisions is expected to radiate thermal dilepton radiation. However, this thermal dilepton radiation interferes with dileptons originating from hadron decays. In the invariant mass region between the $ϕ$ and $J/Ψ$ peak ($1\,$GeV$\lesssim M_{\ell^+ \ell^-} \lesssim 3 \,$GeV) the most substantial background of hadron decays originates from correlated D$\bar{\mathrm{D}}$-meson decays. We evaluate this background using a Langevin simulation for charm quarks. As background medium we utilize the well-tested UrQMD-hybrid model. The required drag and diffusion coefficients are taken from a resonance approach. The decoupling of the charm quarks from the hot medium is performed at a temperature of $130\,$MeV and as hadronization mechanism a coalescence approach is chosen. This model for charm quark interactions with the medium has already been successfully applied to the study of the medium modification and the elliptic flow at FAIR, RHIC and LHC energies. In this proceeding we present our results for the dilepton radiation from correlated D$\bar{\mathrm{D}}$ decays at RHIC energy in comparison to PHENIX measurements in the invariant mass range between 1 and 3 GeV using different interaction scenarios. These results can be utilized to estimate the thermal QGP radiation.

hep-ph↗

arXiv2013

Dileptons from correlated D- and $\bar{\text{D}}$-meson decays in the invariant mass range of the QGP thermal radiation using the UrQMD hybrid model

Relativistic heavy-ion collisions produce a hot and dense thermalized medium, that is expected to emit thermal radiation in form of dileptons. These dileptons are not affected by the strong force and are therefore a clean probe for the possible creation of a Quark Gluon Plasma (QGP). However, electroweak decays of open-charm mesons are expected to induce a substantial background in the invariant mass region between the $ϕ$ and $J/Ψ$ peak ($1\,\text{GeV} \lesssim M_{\ell^+ \ell^-} \lesssim 3 \,\text{GeV}$) of the thermal QGP radiation. To evaluate this background radiation we apply a Langevin approach for the transport of charm quarks in the UrQMD (hydrodynamics + Boltzmann) hybrid model. Due to the inclusion of event-by-event fluctuations and a full (3+1)-dimensional hydrodynamic evolution, the UrQMD hybrid approach provides a more realistic model for the evolution of the matter produced in heavy ion collisions as compared to simple homogeneous fireball expansions usually employed before. As drag and diffusion coefficients we use a resonance approach for elastic heavy-quark scattering and assume a decoupling temperature of the charm quarks from the hot medium of $130\,\text{MeV}$. For the hadronization of the charm quarks we employ a coalescence approach at the decoupling temperature from the medium. In this letter we present our calculations of the D-meson correlations and the invariant mass spectra of the dilepton decays in heavy-ion collisions at FAIR, RHIC, and LHC energies using different interaction scenarios.

hep-ph↗

arXiv2013

Elliptic flow and nuclear modification factors of D-mesons at FAIR in a Hybrid-Langevin approach

The Compressed Baryonic Matter (CBM) experiment at the Facility for Anti-proton and Ion Research (FAIR) will provide new possibilities for charm-quark ($D$-meson) observables in heavy-ion collisions at low collision energies and high baryon densities. To predict the collective flow and nuclear modification factors of charm quarks in this environment, we apply a Langevin approach for the transport of charm quarks in the UrQMD (hydrodynamics + Boltzmann) hybrid model. Due to the inclusion of event-by-event fluctuations and a full (3+1) dimensional hydrodynamical evolution, the UrQMD hybrid approach provides a realistic evolution of the matter produced in heavy-ion collisions. As drag and diffusion coefficients we use a resonance approach for elastic heavy-quark scattering and assume a decoupling temperature of the charm quarks from the hot medium of $130\, \MeV$. Hadronization of the charm quarks to $D$-mesons by coalescence is included. Since the initial charm-quark distribution at FAIR is unknown, we utilize two different initial charm-quark distributions in our approach to estimate the uncertainty of these predictions. We present calculations of the nuclear modification factor, $R_{AA}$, as well as for the elliptic flow, $v_2$, in Pb+Pb collisions at $E_{lab}=25\,\text{AGeV}$. The different medium modifications of $D$-mesons and $\bar{D}$-mesons at high baryon-chemical potential are explored by modified drag- and diffusion-coefficients using the corresponding fugacity factor. Here we find a considerably larger medium modification for $\bar{\mathrm{D}}$- than for $D$-mesons.

hep-ph↗

arXiv2013

Possibility for $J/Ψ$ suppression in high multiplicity proton-proton collisions at $\sqrt{s_{NN}}=7\,$TeV

We study $J/Ψ$ absorption in high multiplicity proton-proton (pp) collisions at $\sqrt {s_{NN}}=7\,$TeV. We predict a modification of the $J/Ψ$-yield within the UrQMD transport approach, where explicit interactions of the $J/Ψ$ with the surrounding comovers and a prehadronic phase with adjusted cross sections and $J/Ψ$ melting is included. We present the analog of the nuclear modification factor in pp collisions at various charged particle multiplicities. It turns out that $J/Ψ$-Mesons may be suppressed towards higher particle multiplicities in pp collisions at LHC energies.

hep-ph↗

arXiv2012

Heavy quark transport at RHIC and LHC

We calculate the heavy quark evolution in heavy ion collisions and show results for the elliptic flow $v_2$ as well as the nuclear modification factor $R_{AA}$ at RHIC and LHC energies. For the calculation we implement a Langevin approach for the transport of heavy quarks in the UrQMD (hydrodynamics + Boltzmann) hybrid model. As drag and diffusion coefficients we use a Resonance approach for elastic heavy-quark scattering and assume a decoupling temperature of the charm quarks from the hot medium of $130\,\text{MeV}$. At RHIC energies we use a coalescence approach at the decoupling temperature for the hadronization of the heavy quarks to D-mesons and B-mesons and a sub-following decay to heavy flavor electrons using PYTHIA. At LHC we use an additional fragmentation mechanism to account for the higher transverse momenta reached at higher collision energies.

hep-ph↗

arXiv2012

Heavy quark transport in heavy ion collisions at RHIC and LHC within the UrQMD transport model

We have implemented a Langevin approach for the transport of heavy quarks in the UrQMD hybrid model. The UrQMD hybrid approach provides a realistic description of the background medium for the evolution of relativistic heavy ion collisions. We have used two different sets of drag and diffusion coefficients, one based on a $T$-Matrix approach and one based on a resonance model for the elastic scattering of heavy quarks within the medium. In case of the resonance model we have investigated the effects of different decoupling temperatures of the heavy quarks from the medium, ranging between $130\,\text{MeV}$ and $180\,\text{MeV}$. We present calculations of the nuclear modification factor $R_{AA}$, as well as of the elliptic flow $v_2$ in Au+Au collisions at $\sqrt{s_{NN}}=200\,\text{GeV}$ and Pb+Pb collisions at $\sqrt{s_{NN}}=2.76\,\text{TeV}$. To make our results comparable to experimental data at RHIC and LHC we have implemented a Peterson fragmentation and a quark coalescence approach followed by the semileptonic decay of the D- and B-mesons to electrons. We find that our results strongly depend on the decoupling temperature and the hadronization mechanism. At a decoupling temperature of $130\,\text{MeV}$ we reach a good agreement with the measurements at both, RHIC and LHC energies, simultaneously for the elliptic flow $v_2$ and the nuclear modification factor $R_{AA}$.

hep-ph↗