arXiv subjects

Frank-Peter Schilling

Publications and source records attributed to Frank-Peter Schilling.

16 recordsLinked to original sources

arXiv2026

Dual-domain U-Nets with embedded back projection operators for motion-resolved 4D CBCT reconstruction

Four-dimensional cone beam CT (4D CBCT) is important for image-guided radiation therapy of thoracic cancers, but its use is limited by long scan times, causing high patient dose and motion/sparse-sampling artifacts. We propose a deep learning method for motion-resolved 4D CBCT reconstruction from conventional free-breathing scans, without a respiratory signal or explicit projection binning. Our CNN takes free-breathing 3D CBCT projections as input and predicts a static volume at maximum inhalation plus ten displacement vector fields (DVFs) spanning a breathing cycle. The network extends U-Net: the encoder acts on filtered projection stacks, the decoder acts in the volume domain, and skip connections are replaced with non-trainable back-projection functions at multiple resolutions to transfer features between domains. The model is trained on simulated CBCT scans and evaluated on 11 unseen simulated patients and 13 clinical free-breathing scans. Two additional models (60 s and 6 s scans) were evaluated by clinical experts on three and two scans, comparing single phases of our 4D reconstruction to reference 3D SART-TV images for tumor and esophagus visibility. Experts preferred our method for tumor visibility (59% vs. 36% no preference, 5% reference) and esophagus visibility (47% vs. 42%, 11%). On simulated data, image quality matched SART-TV (mean RMSE: -1.19 HU, PSNR: +0.09 dB, SSIM: -0.009) while enabling 4D reconstruction. On clinical scans, our method showed sharper dynamic structures (e.g., diaphragm) and fewer motion streak artifacts than traditional reconstruction. This non-patient-specific CNN predicts static volumes and full 4D respiratory motion models from a single free-breathing scan, without a respiratory surrogate or projection binning, reducing motion artifacts while adding motion-modeling capability.

cs.CV↗

arXiv2026

Projected Energy Matching for Generative 3D Priors

Energy Matching has emerged as a powerful generative framework that combines flow model efficiency with the explicit likelihood of Energy-Based Models (EBMs) via a single, time-independent scalar potential. However, directly training this potential on high-dimensional 3D data remains computationally challenging. While distilling a pre-trained flow model circumvents some of the initial training costs, we demonstrate that velocity fields inevitably contain non-conservative rotational artifacts (curl). Forcing a strictly conservative scalar potential to match this unconstrained field creates a "structural conflict", which degrades generation quality and mode coverage. To solve this, we propose Projected Energy Matching, a scalable framework that resolves these structural and computational bottlenecks. We introduce Helmholtz Distillation, a structural relaxation that leverages a Hutchinson trace estimator to explicitly absorb rotational noise into an auxiliary residual network. We subsequently refine this landscape using Negative Caching, a memory-efficient strategy that reuses negative samples across micro-batches, rendering sampling tractable during contrastive training with gradient accumulation. We deploy our method as an unconditional prior for real-world medical CT inverse problems, specifically sparse-view reconstruction. Ultimately, our amortized pipeline reduces total compute to a small fraction of that required by standard energy matching, while achieving high-fidelity reconstructions and successfully resolving severe measurement artifacts.

eess.IV↗

arXiv2026

The Role of Artificial Intelligence in the SKA Era

The Square Kilometre Array Observatory (SKAO) will usher in an era of unprecedented data complexity and scientific opportunity in radio astronomy, producing petabyte-scale datasets and terabit-per-second streams that challenge traditional analysis paradigms. Artificial Intelligence (AI) stands at the forefront of this transformation, offering scalable, adaptive solutions to the most pressing problems in radio astronomy and astrophysics. This chapter explores the pivotal role of AI in the SKA era, from real-time operations to scientific discovery. We examine how deep learning models enable automated source detection, radio-frequency interference mitigation, anomaly detection, and parameter inference, while generative approaches accelerate sky simulations, calibration, and imaging. Reinforcement learning promises dynamic scheduling and autonomous system control, and federated learning could address the distributed nature of SKA data. Beyond performance, we emphasize the necessity of explainability, uncertainty quantification, and physics-informed inductive biases to ensure scientific integrity. By mapping SKAO's core challenges - data volume, complexity, and interpretability - onto modern AI methodologies, we review how deep learning, self-supervised frameworks, and probabilistic models can unlock new frontiers in cosmology, galaxy evolution, and time-domain astrophysics. AI is not merely an automation tool for coping with scale. It is a catalyst for discovery, redefining how we observe, model, and understand the Universe.

astro-ph.IM↗

arXiv2026

Optimization of Deep Learning Models for Radio Galaxy Classification

Modern radio telescope surveys, capable of detecting billions of galaxies in wide-field surveys, have made manual morphological classification impracticable. This applies in particular when the Square Kilometre Array Observatory (SKAO) becomes operable in 2027, which is expected to close an important gap in our understanding of the Epoch of Reionization (EoR) and other areas of astrophysics. To this end, foreground objects, contaminants of the 21-cm signal, need to be identified and subtracted. Source finding and identification is thus an important albeit challenging task. We investigate the ability of AI and deep learning (DL) methods that have been previously trained on other data domains to localize and classify radio galaxies with minimal changes to their architectures. Various well-known pretrained neural network architectures for image classification and object detection are trained and fine-tuned and their performance is evaluated on a public radio galaxy dataset derived from the Radio Galaxy Zoo. A comparison between convolutional neural network (CNN)- and transformer-based algorithms is performed. The best performing architecture is systematically optimized and an uncertainty estimation is performed by means of an ensemble analysis. Radio source classification performance nearly comparable to the current leading customized models can be obtained using existing standard pretrained DL architectures, without modification and increase in complexity of the model architectures but rather adaptation of the data, by combining various transformations on replicated image channels. Using an ensemble of models can also further improve performance to over 90% accuracy, on par with top-performing models in the literature. The results can be transferred to other survey data, e.g. from the Murchison Wide-field Array (MWA), and in the future be used to study the EoR with the SKAO.

astro-ph.GA↗

arXiv2025

Galactic Alchemy: Deep Learning Map-to-Map Translation in Hydrodynamical Simulations

We present the first systematic study of multi-domain map-to-map translation in galaxy formation simulations, leveraging deep generative models to predict diverse galactic properties. Using high-resolution magneto-hydrodynamical simulation data, we compare conditional generative adversarial networks and diffusion models under unified preprocessing and evaluation, optimizing architectures and attention mechanisms for physical fidelity on galactic scales. Our approach jointly addresses seven astrophysical domains - including dark matter, gas, neutral hydrogen, stellar mass, temperature, and magnetic field strength - while introducing physics-aware evaluation metrics that quantify structural realism beyond standard computer vision measures. We demonstrate that translation difficulty correlates with physical coupling, achieving near-perfect fidelity for mappings from gas to dark matter and mappings involving astro-chemical components such as total gas to HI content, while identifying fundamental challenges in weakly constrained tasks such as gas to stellar mass mappings. Our results establish GAN-based models as competitive counterparts to state-of-the-art diffusion approaches at a fraction of the computational cost (in training and inference), paving the way for scalable, physics-aware generative frameworks for forward modelling and observational reconstruction in the SKA era.

astro-ph.GA↗

arXiv2025

MInDI-3D: Iterative Deep Learning in 3D for Sparse-view Cone Beam Computed Tomography

We present MInDI-3D (Medical Inversion by Direct Iteration in 3D), the first 3D conditional diffusion-based model for real-world sparse-view Cone Beam Computed Tomography (CBCT) artefact removal, aiming to reduce imaging radiation exposure. A key contribution is extending the "InDI" concept from 2D to a full 3D volumetric approach for medical images, implementing an iterative denoising process that refines the CBCT volume directly from sparse-view input. A further contribution is the generation of a large pseudo-CBCT dataset (16,182) from chest CT volumes of the CT-RATE public dataset to robustly train MInDI-3D. We performed a comprehensive evaluation, including quantitative metrics, scalability analysis, generalisation tests, and a clinical assessment by 11 clinicians. Our results show MInDI-3D's effectiveness, achieving a 12.96 (6.10) dB PSNR gain over uncorrected scans with only 50 projections on the CT-RATE pseudo-CBCT (independent real-world) test set and enabling an 8x reduction in imaging radiation exposure. We demonstrate its scalability by showing that performance improves with more training data. Importantly, MInDI-3D matches the performance of a 3D U-Net on real-world scans from 16 cancer patients across distortion and task-based metrics. It also generalises to new CBCT scanner geometries. Clinicians rated our model as sufficient for patient positioning across all anatomical sites and found it preserved lung tumour boundaries well.

cs.CV↗

arXiv2024

Artifact Reduction in 3D and 4D Cone-beam Computed Tomography Images with Deep Learning -- A Review

Deep learning based approaches have been used to improve image quality in cone-beam computed tomography (CBCT), a medical imaging technique often used in applications such as image-guided radiation therapy, implant dentistry or orthopaedics. In particular, while deep learning methods have been applied to reduce various types of CBCT image artifacts arising from motion, metal objects, or low-dose acquisition, a comprehensive review summarizing the successes and shortcomings of these approaches, with a primary focus on the type of artifacts rather than the architecture of neural networks, is lacking in the literature. In this review, the data generation and simulation pipelines, and artifact reduction techniques are specifically investigated for each type of artifact. We provide an overview of deep learning techniques that have successfully been shown to reduce artifacts in 3D, as well as in time-resolved (4D) CBCT through the use of projection- and/or volume-domain optimizations, or by introducing neural networks directly within the CBCT reconstruction algorithms. Research gaps are identified to suggest avenues for future exploration. One of the key findings of this work is an observed trend towards the use of generative models including GANs and score-based or diffusion models, accompanied with the need for more diverse and open training datasets and simulations.

cs.CV↗

arXiv2013

Top Quark Physics at the LHC: A Review of the First Two Years

This review summarizes the highlights in the area of top quark physics obtained with the two general purpose detectors ATLAS and CMS during the first two years of operation of the Large Hadron Collider LHC. It covers the 2010 and 2011 data taking periods, where the LHC provided pp collisions at a center-of-mass energy of sqrt(s)=7 TeV. Measurements are presented of the total and differential top quark pair production cross section in many different channels, the top quark mass and various other properties of the top quark and its interactions, for instance the charge asymmetry. Measurements of single top quark production and various searches for new physics involving top quarks are also discussed. The already very precise experimental data are in good agreement with the standard model.

hep-ex↗

arXiv2013

Measurements of the Top Quark Pair-Production Cross Section

Measurements of the inclusive and differential cross section for the production of top quark pairs in proton-(anti)proton collision at center-of-mass energies of 1.96, 7.0 and 8.0 TeV are presented and compared with the latest theory predictions and Monte-Carlo models. In addition, first measurements of the production of top quark pairs in association with additional jets or with a boson are highlighted. All measurements are in good agreement with the Standard Model.

hep-ex↗

arXiv2010

Top Quark Studies with the first CMS Data

Studies are presented of the selection of events consistent with top quark pair production in data recorded by the CMS detector at the LHC, corresponding to an integrated luminosity of 0.84+/-0.09 1/pb and at center-of-mass energy sqrt{s}=7 TeV. Results are presented for the lepton+jets as well as dilepton channels. Event yields in data are compared to those in simulation, and several background processes are estimated using data-driven techniques. The observed yields of top-antitop candidate events are roughly consistent with the Standard Model.

hep-ex↗

arXiv2009

QCD and Top Quark Physics at the LHC

The expected performance of the ATLAS and CMS detectors at the Large Hadron Collider (LHC) in QCD and top quark measurements is discussed, with a focus on the early data taking phase. Such processes are amongst the primary backgrounds in the searches for new physics, and thus must be understood very well before discoveries can be made. In addition, they serve as useful detector calibration candles.

hep-ex↗

arXiv2007

Track Based Alignment in CMS

The strategy for track based alignment of the CMS tracking and muon detectors is presented. After an overview over the used data samples, the general alignment strategy is presented, with a focus of the procedures envisaged at the start of data taking in 2008. The three currently used alignment algorithms are discussed and first results on their application to the CMS tracker are presented, as well as studies on the alignment of the Muon detector with tracks.

physics.ins-det↗

arXiv2005

HERA Diffractive Structure Function Data and Parton Distributions

Recent diffractive structure function measurements by the H1 and ZEUS experiments at HERA are reviewed. Various data sets, obtained using systematically different selection and reconstruction methods, are compared. NLO DGLAP QCD fits are performed to the most precise H1 and ZEUS data and diffractive parton densities are obtained in each case. Differences between the Q^2 dependences of the H1 and ZEUS data are reflected as differences between the diffractive gluon densities.

hep-ex↗

arXiv2005

Prospects for Measuring B_s->mu mu with the CMS Detector

The flavor-changing neutral current decay B_s->mu mu is highly suppressed in the standard model, but its branching fraction of 3.4*10^-9 could be significantly enhanced through contributions from new physics. At the LHC, this rare decay could be observed for the first time. In this contribution, the prospects for measuring B_s->mu mu with the CMS detector are presented. In particular, some aspects of the experimental setup, the first and high level trigger selections, and the offline analysis are discussed.

hep-ex↗

arXiv2003

Diffractive Final States with the H1 Detector at HERA

Recent measurements of diffractive dijet and charm quark production in electron-proton collisions using the H1 detector at HERA are presented, where the exchanged photon is either almost real or highly virtual. The data are compared with leading and next-to-leading order QCD calculations based on the diffractive parton distributions obtained from a recent DGLAP QCD analysis of H1 inclusive diffractive deep-inelastic scattering data, thus testing QCD factorization in diffractive ep interactions.

hep-ex↗

arXiv2002

NLO QCD Fit to H1 Diffractive DIS Data

A new NLO DGLAP QCD fit to recent inclusive diffractive DIS data from the H1 collaboration is presented. Diffractive parton distributions are extracted, including their experimental and theoretical uncertainties. The parton distributions are used for comparisons with recent diffractive final state data from HERA and the TEVATRON.

hep-ex↗