Bayesian model selection and misspecification testing in imaging inverse problems only from noisy and partial measurements

Tom Sprunck, Marcelo Pereyra, Tobı́as I. Liaudat
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:115394-115419, 2026.

Abstract

Modern imaging techniques heavily rely on Bayesian statistical models to address difficult image reconstruction and restoration tasks. This paper addresses the objective evaluation of such models in settings where ground truth is unavailable, with a focus on model selection and misspecification diagnosis. Existing unsupervised model evaluation methods are often unsuitable for computational imaging due to their high computational cost and incompatibility with modern image priors defined implicitly via machine learning models. We herein propose a general methodology for unsupervised model selection and misspecification detection in Bayesian imaging sciences, based on a novel combination of Bayesian cross-validation and data fission, a randomized measurement splitting technique. The approach is compatible with any Bayesian imaging sampler, including diffusion and plug-and-play samplers. We demonstrate the methodology through experiments involving various scoring rules and types of model misspecification, where we achieve excellent selection and detection accuracy with a low computational cost.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-sprunck26a, title = {{B}ayesian model selection and misspecification testing in imaging inverse problems only from noisy and partial measurements}, author = {Sprunck, Tom and Pereyra, Marcelo and Liaudat, Tob\'{\i}as I.}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {115394--115419}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/sprunck26a/sprunck26a.pdf}, url = {https://proceedings.mlr.press/v306/sprunck26a.html}, abstract = {Modern imaging techniques heavily rely on Bayesian statistical models to address difficult image reconstruction and restoration tasks. This paper addresses the objective evaluation of such models in settings where ground truth is unavailable, with a focus on model selection and misspecification diagnosis. Existing unsupervised model evaluation methods are often unsuitable for computational imaging due to their high computational cost and incompatibility with modern image priors defined implicitly via machine learning models. We herein propose a general methodology for unsupervised model selection and misspecification detection in Bayesian imaging sciences, based on a novel combination of Bayesian cross-validation and data fission, a randomized measurement splitting technique. The approach is compatible with any Bayesian imaging sampler, including diffusion and plug-and-play samplers. We demonstrate the methodology through experiments involving various scoring rules and types of model misspecification, where we achieve excellent selection and detection accuracy with a low computational cost.} }
Endnote
%0 Conference Paper %T Bayesian model selection and misspecification testing in imaging inverse problems only from noisy and partial measurements %A Tom Sprunck %A Marcelo Pereyra %A Tobı́as I. Liaudat %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-sprunck26a %I PMLR %P 115394--115419 %U https://proceedings.mlr.press/v306/sprunck26a.html %V 306 %X Modern imaging techniques heavily rely on Bayesian statistical models to address difficult image reconstruction and restoration tasks. This paper addresses the objective evaluation of such models in settings where ground truth is unavailable, with a focus on model selection and misspecification diagnosis. Existing unsupervised model evaluation methods are often unsuitable for computational imaging due to their high computational cost and incompatibility with modern image priors defined implicitly via machine learning models. We herein propose a general methodology for unsupervised model selection and misspecification detection in Bayesian imaging sciences, based on a novel combination of Bayesian cross-validation and data fission, a randomized measurement splitting technique. The approach is compatible with any Bayesian imaging sampler, including diffusion and plug-and-play samplers. We demonstrate the methodology through experiments involving various scoring rules and types of model misspecification, where we achieve excellent selection and detection accuracy with a low computational cost.
APA
Sprunck, T., Pereyra, M. & Liaudat, T.I.. (2026). Bayesian model selection and misspecification testing in imaging inverse problems only from noisy and partial measurements. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:115394-115419 Available from https://proceedings.mlr.press/v306/sprunck26a.html.

Related Material