M8. Inverse Problems with Data-Driven Methods and Deep Learning
In recent years, the volume and complexity of various data obtained by modern applications has been steadily increasing. Extracting consistent scientific information from these large datasets remains an open challenge for the community, and data-driven approaches such as deep learning have quickly emerged as a potentially powerful solution to some long-term problems. In this context, robust mathematical inversion algorithms, combined with new data science techniques, deliver state-of-the-art results across a wide range of inverse problems.
This mini-symposium aims at bringing together experts in data-driven methods and deep learning for inverse problems and provides an overview of learned image reconstruction approaches, mathematical insights, and real-world applications.
Organizers:
Tatiana Bubba, University of Bath, UK, This email address is being protected from spambots. You need JavaScript enabled to view it.
Andreas Hauptmann, University of Oulu, Finland, This email address is being protected from spambots. You need JavaScript enabled to view it.
Luca Ratti, University of Bologna, Italy, This email address is being protected from spambots. You need JavaScript enabled to view it.
Invited Speakers:
Tan Bui-Thanh, The University of Texas, USA, This email address is being protected from spambots. You need JavaScript enabled to view it.
Model-constrained uncertainty quantification for scientific deep learning of inverse problems
Marcello Carioni, University of Twente, The Netherlands, This email address is being protected from spambots. You need JavaScript enabled to view it.
Optimal transport methods for inverse problems regularization
Margaret Duff, STFC – Rutherford Appleton Laboratories, UK, This email address is being protected from spambots. You need JavaScript enabled to view it.
VAEs with structured image covariance as priors to inverse imaging problems
Felix Herrmann, Georgia Institute of Technology, USA, This email address is being protected from spambots. You need JavaScript enabled to view it.
Neural wave-based imaging with amortized uncertainty quantification
Johannes Hertrich, University College London, UK, This email address is being protected from spambots. You need JavaScript enabled to view it.
Generative sliced MMD flows for posterior sampling in Bayesian Inverse Problems
Erich Kobler, University of Bonn, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Learning gradually non-convex image priors using score matching
Andreas Kofler, Physikalisch-Technische Bundesanstalt (PTB), Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Learning spatio-temporal regularization parameter maps for TV-minimization-based image reconstruction
Rémi Laumont, Technical University of Denmark, Denmark, This email address is being protected from spambots. You need JavaScript enabled to view it.
Bayesian computation with Plug-and-Play (PnP) priors for inverse problems in imaging sciences
Dong Liu, University of Science and Technology of China, China, This email address is being protected from spambots. You need JavaScript enabled to view it.
Unsupervised neural networks for image reconstruction
Subhadip Mukherjee, IIT Kharagpur, India, This email address is being protected from spambots. You need JavaScript enabled to view it.
Provably convergent plug-and-play quasi-Newton methods
Sebastian Neumayer, TU Chemnitz, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Learning spatially-adaptive regularization
Thomas Pinetz, Bonn University, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Blind single image super-resolution via iterated shared prior learning
Luca Ratti, University of Bologna, Italy, This email address is being protected from spambots. You need JavaScript enabled to view it.
Learned reconstruction methods for inverse problems: sample error estimates
Robert Scheichl, University of Heidelberg, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Deep Inverse Rosenblatt Transport for Bayesian Inverse Problems
Aslam Shaikh, Aalto University, Finland, This email address is being protected from spambots. You need JavaScript enabled to view it.
Specimen reconstruction in atom probe tomography as an inverse problem
Johan S. Wind, University of Oslo, Norway, This email address is being protected from spambots. You need JavaScript enabled to view it.
Implicit regularization in AI meets generalized hardness of approximation: Sharp results for diagonal linear networks