Self-Supervised Uncertainty Estimation For Super-Resolution of Satellite Images

Zhe Zheng, Valéry Dewil, Pablo Arias
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:8238-8252, 2026.

Abstract

Super-resolution (SR) of satellite imagery is challenging due to the lack of paired low-/high-resolution data. Recent self-supervised SR methods overcome this limitation by exploiting the temporal redundancy in burst observations, but they lack a mechanism to quantify uncertainty in the reconstruction. In this work, we introduce a novel self-supervised loss that allows to estimate uncertainty in image super-resolution without ever accessing the ground-truth high-resolution data. We adopt a decision-theoretic perspective and show that minimizing the corresponding {Bayesian} risk yields the posterior mean and variance as optimal estimators. We validate our approach on a synthetic dataset with both white and signal dependent noise and demonstrate that it produces calibrated uncertainty estimates comparable to supervised methods. We further apply our method on real SkySat satellite data and validate its performance through an unsupervised coverage test. Our work bridges self-supervised restoration with uncertainty quantification, making a practical framework for uncertainty-aware image reconstruction.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-zheng26a, title = {Self-Supervised Uncertainty Estimation For Super-Resolution of Satellite Images}, author = {Zheng, Zhe and Dewil, Val\'{e}ry and Arias, Pablo}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {8238--8252}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/zheng26a/zheng26a.pdf}, url = {https://proceedings.mlr.press/v337/zheng26a.html}, abstract = {Super-resolution (SR) of satellite imagery is challenging due to the lack of paired low-/high-resolution data. Recent self-supervised SR methods overcome this limitation by exploiting the temporal redundancy in burst observations, but they lack a mechanism to quantify uncertainty in the reconstruction. In this work, we introduce a novel self-supervised loss that allows to estimate uncertainty in image super-resolution without ever accessing the ground-truth high-resolution data. We adopt a decision-theoretic perspective and show that minimizing the corresponding {Bayesian} risk yields the posterior mean and variance as optimal estimators. We validate our approach on a synthetic dataset with both white and signal dependent noise and demonstrate that it produces calibrated uncertainty estimates comparable to supervised methods. We further apply our method on real SkySat satellite data and validate its performance through an unsupervised coverage test. Our work bridges self-supervised restoration with uncertainty quantification, making a practical framework for uncertainty-aware image reconstruction.} }
Endnote
%0 Conference Paper %T Self-Supervised Uncertainty Estimation For Super-Resolution of Satellite Images %A Zhe Zheng %A Valéry Dewil %A Pablo Arias %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-zheng26a %I PMLR %P 8238--8252 %U https://proceedings.mlr.press/v337/zheng26a.html %V 337 %X Super-resolution (SR) of satellite imagery is challenging due to the lack of paired low-/high-resolution data. Recent self-supervised SR methods overcome this limitation by exploiting the temporal redundancy in burst observations, but they lack a mechanism to quantify uncertainty in the reconstruction. In this work, we introduce a novel self-supervised loss that allows to estimate uncertainty in image super-resolution without ever accessing the ground-truth high-resolution data. We adopt a decision-theoretic perspective and show that minimizing the corresponding {Bayesian} risk yields the posterior mean and variance as optimal estimators. We validate our approach on a synthetic dataset with both white and signal dependent noise and demonstrate that it produces calibrated uncertainty estimates comparable to supervised methods. We further apply our method on real SkySat satellite data and validate its performance through an unsupervised coverage test. Our work bridges self-supervised restoration with uncertainty quantification, making a practical framework for uncertainty-aware image reconstruction.
APA
Zheng, Z., Dewil, V. & Arias, P.. (2026). Self-Supervised Uncertainty Estimation For Super-Resolution of Satellite Images. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:8238-8252 Available from https://proceedings.mlr.press/v337/zheng26a.html.

Related Material