Publications | Dr Teresa Klatzer

2026

2025

Efficient Bayesian Computation Using Plug-and-Play Priors for Poisson Inverse Problems

Teresa Klatzer, Savvas Melidonis, Marcelo Pereyra, and 1 more author

SIAM Journal on Imaging Sciences (in press), 2025

Abs Bib

This paper introduces a novel plug-and-play (PnP) Langevin sampling methodology for Bayesian inference in low-photon Poisson imaging problems, a challenging class of problems with significant applications in astronomy, medicine, and biology. PnP Langevin sampling algorithms offer a powerful framework for Bayesian image restoration, enabling accurate point estimation as well as advanced inference tasks, including uncertainty quantification and visualization analyses, and empirical Bayesian inference for automatic model parameter tuning. However, existing PnP Langevin algorithms are not well-suited for low-photon Poisson imaging due to high solution uncertainty and poor regularity properties, such as exploding gradients and non-negativity constraints. To address these challenges, we propose two strategies for extending Langevin PnP sampling to Poisson imaging models: (i) an accelerated PnP Langevin method that incorporates boundary reflections and a Poisson likelihood approximation and (ii) a mirror sampling algorithm that leverages a Riemannian geometry to handle the constraints and the poor regularity of the likelihood without approximations. The effectiveness of these approaches is demonstrated through extensive numerical experiments and comparisons with state-of-the-art methods.
@article{klatzer2025efficientbayesiancomputationusing, title = {Efficient Bayesian Computation Using Plug-and-Play Priors for Poisson Inverse Problems}, author = {Klatzer, Teresa and Melidonis, Savvas and Pereyra, Marcelo and Zygalakis, Konstantinos C.}, year = {2025}, eprint = {2503.16222}, journal = {SIAM Journal on Imaging Sciences (in press)}, archiveprefix = {arXiv}, primaryclass = {stat.CO}, }

2024

Accelerated Bayesian imaging by relaxed proximal-point Langevin sampling

Teresa Klatzer, Paul Dobson, Yoann Altmann, and 3 more authors

SIAM Journal on Imaging Sciences, 2024

Abs arXiv Bib Code Slides

This paper presents a new accelerated proximal Markov chain Monte Carlo methodology to perform Bayesian inference in imaging inverse problems with an underlying convex geometry. The proposed strategy takes the form of a stochastic relaxed proximal-point iteration that admits two complementary interpretations. For models that are smooth or regularised by Moreau-Yosida smoothing, the algorithm is equivalent to an implicit midpoint discretisation of an overdamped Langevin diffusion targeting the posterior distribution of interest. This discretisation is asymptotically unbiased for Gaussian targets and shown to converge in an accelerated manner for any target that is κ-strongly log-concave (i.e., requiring in the order of √κ iterations to converge, similarly to accelerated optimisation schemes), comparing favorably to [M. Pereyra, L. Vargas Mieles, K.C. Zygalakis, SIAM J. Imaging Sciences, 13,2 (2020), pp. 905-935] which is only provably accelerated for Gaussian targets and has bias. For models that are not smooth, the algorithm is equivalent to a Leimkuhler-Matthews discretisation of a Langevin diffusion targeting a Moreau-Yosida approximation of the posterior distribution of interest, and hence achieves a significantly lower bias than conventional unadjusted Langevin strategies based on the Euler-Maruyama discretisation. For targets that are κ-strongly log-concave, the provided non-asymptotic convergence analysis also identifies the optimal time step which maximizes the convergence speed. The proposed methodology is demonstrated through a range of experiments related to image deconvolution with Gaussian and Poisson noise, with assumption-driven and data-driven convex priors.
@article{klatzer2024accelerated, title = {Accelerated Bayesian imaging by relaxed proximal-point Langevin sampling}, author = {Klatzer, Teresa and Dobson, Paul and Altmann, Yoann and Pereyra, Marcelo and Sanz-Serna, Jes{\'u}s Mar{\'\i}a and Zygalakis, Konstantinos C}, journal = {SIAM Journal on Imaging Sciences}, volume = {17}, number = {2}, pages = {1078--1117}, year = {2024}, publisher = {SIAM}, doi = {10.1137/23M1594832}, }

2018

Learning a variational network for reconstruction of accelerated MRI data

Kerstin Hammernik, Teresa Klatzer, Erich Kobler, and 4 more authors

In , 2018

Abs arXiv Bib

Purpose To allow fast and high-quality reconstruction of clinical accelerated multi-coil MR data by learning a variational network that combines the mathematical structure of variational models with deep learning. Theory and Methods Generalized compressed sensing reconstruction formulated as a variational model is embedded in an unrolled gradient descent scheme. All parameters of this formulation, including the prior model defined by filter kernels and activation functions as well as the data term weights, are learned during an offline training procedure. The learned model can then be applied online to previously unseen data. Results The variational network approach is evaluated on a clinical knee imaging protocol for different acceleration factors and sampling patterns using retrospectively and prospectively undersampled data. The variational network reconstructions outperform standard reconstruction algorithms, verified by quantitative error measures and a clinical reader study for regular sampling and acceleration factor 4. Conclusion Variational network reconstructions preserve the natural appearance of MR images as well as pathologies that were not included in the training data set. Due to its high computational performance, that is, reconstruction time of 193 ms on a single graphics card, and the omission of parameter tuning once the network is trained, this new approach to image reconstruction can easily be integrated into clinical workflow. Magn Reson Med 79:3055–3071, 2018. © 2017 International Society for Magnetic Resonance in Medicine.
@inproceedings{DBLP:journals/corr/HammernikKKRSPK17, author = {Hammernik, Kerstin and Klatzer, Teresa and Kobler, Erich and Recht, Michael P. and Sodickson, Daniel K. and Pock, Thomas and Knoll, Florian}, title = {Learning a variational network for reconstruction of accelerated MRI data}, journal = {Magnetic Resonance in Medicine}, volume = {79}, number = {6}, pages = {3055-3071}, keywords = {variational network, deep learning, accelerated MRI, parallel imaging, compressed sensing, image reconstruction}, doi = {10.1002/mrm.26977}, url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/mrm.26977}, year = {2018}, }
Variational Networks for Joint Image Reconstruction and Classification of Tumor Immune Cell Interactions in Melanoma Tissue Sections

Alexander. Effland, Michael Hölzel, Teresa Klatzer, and 5 more authors

In Bildverarbeitung in der Medizin 2018, 2018

2017

Variational Networks: Connecting Variational Methods and Deep Learning

Erich Kobler, Teresa Klatzer, Kerstin Hammernik, and 1 more author

In Pattern Recognition. GCPR 2017, 2017
Trainable Regularization for Multi-frame Superresolution

Teresa Klatzer, Daniel Soukup, Erich Kobler, and 2 more authors

In Pattern Recognition, 2017

Abs

In this paper, we present a novel method for multi-frame superresolution (SR). Our main goal is to improve the spatial resolution of a multi-line scan camera for an industrial inspection task. High resolution output images are reconstructed using our proposed SR algorithm for multi-channel data, which is based on the trainable reaction-diffusion model. As this is a supervised learning approach, we simulate ground truth data for a real imaging scenario. We show that learning a regularizer for the SR problem improves the reconstruction results compared to an iterative reconstruction algorithm using TV or TGV regularization. We test the learned regularizer, trained on simulated data, on images acquired with the real camera setup and achieve excellent results.

2016

Learning joint demosaicing and denoising based on sequential energy minimization

Teresa Klatzer, Kerstin Hammernik, Patrick Knobelreiter, and 1 more author

In 2016 IEEE International Conference on Computational Photography (ICCP), May 2016

Abs

Demosaicing is an important first step for color image acquisition. For practical reasons, demosaicing algorithms have to be both efficient and yield high quality results in the presence of noise. The demosaicing problem poses several challenges, e.g. zippering and false color artifacts as well as edge blur. In this work, we introduce a novel learning based method that can overcome these challenges. We formulate demosaicing as an image restoration problem and propose to learn efficient regularization inspired by a variational energy minimization framework that can be trained for different sensor layouts. Our algorithm performs joint demosaicing and denoising in close relation to the real physical mosaicing process on a camera sensor. This is achieved by learning a sequence of energy minimization problems composed of a set of RGB filters and corresponding activation functions. We evaluate our algorithm on the Microsoft Demosaicing data set in terms of peak signal to noise ratio (PSNR) and structured similarity index (SSIM). Our algorithm is highly efficient both in image quality and run time. We achieve an improvement of up to 2.6 dB over recent state-of-the-art algorithms.

2015

Continuous Hyper-parameter Learning for Support Vector Machines

Teresa Klatzer, and Thomas Pock

In Computer Vision Winter Workshop (CVVW), May 2015

Bib

@inproceedings{klatzer2015,
  title = {Continuous Hyper-parameter Learning for Support Vector Machines},
  author = {Klatzer, Teresa and Pock, Thomas},
  year = {2015},
  doi = {10.3217/978-3-85125-388-7},
  language = {English},
  booktitle = {Computer Vision Winter Workshop (CVVW)},
}