Teresa Klatzer

PhD student at University of Edinburgh. she/her. Finishing soon! E-mail me if you want to chat.

portrait_2025.jpg

Bayesian computation and ML in imaging science.

Edinburgh, UK.

🌈

Since 2021, I am working on my PhD in Applied and Computational Mathematics on Bayesian computation for imaging science. My supervisors are Konstantinos Zygalakis, Marcelo Pereyra and Yoann Altmann and I am currently affiliated with the BLOOM project. A list of publications can be found here or on my Google Scholar profile. My published codes can be found here.

My current research interests are at the intersection of Bayesian computation, numerical analysis, inverse problems, and machine learning. I am interested in the mathematical foundations of algorithms and methodology to solve problems in the field of imaging science.

Before joining the project BLOOM, I have been working as a software project manager, product owner and agile coach. I have obtained a professional certification in coaching and counselling in Graz, Austria. I am happy to offer mentoring in mental health topics around science and non-normative walks in life. See my diversity section for more resources.

I have received a MSc in Information and Computer Engineering (Telematics) in 2014 from Graz University of Technology on the topic of “Bi-level Optimization for Support Vector Machines”. Between 2014-2017, I have been working as a research assistant at the Institute for Computer Graphics and Vision in the group of Thomas Pock, working on the project “Deep variational networks for low-level computer vision”. My research topics included the development of machine learning and optimization methods for applications in low-level image processing and computational photography.

News

Mar 20, 2025 I am excited to announce that my latest preprint Efficient Bayesian Computation Using Plug-and-Play Priors for Poisson Inverse Problems is now available on arXiv! 🎉✹
Feb 6, 2025 I feel honoured that my research proposal was selected for the Doctoral Prize Fellowship 🏆 by the Prob_AI hub! See the latest newsletter for the announcement.
Nov 20, 2024 I am excited to attend NeurIPS 2024 in Vancouver, Canada (Dec 10-15) this year! I will present a contributed talk (I am honoured to be 1 out of 4 select presenters!) and a poster about my work at the WiML Workshop. Thanks to the sponsors for the full travel grant! I am presenting my work on “Mirror Langevin Dynamics with Plug-and-Play Priors for Poisson Inverse Problems”.
Mar 10, 2023 Our Python tutorials repository went live! Check it out here. These tutorials are about Bayesian computation and inverse problems in imaging science - to learn about (accelerated) sampling for uncertainty quantification and other advanced inferences.

Selected publications

  1. 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
  2. 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
  3. Continuous Hyper-parameter Learning for Support Vector Machines
    Teresa Klatzer, and Thomas Pock
    In Computer Vision Winter Workshop (CVVW), May 2015