Exploring the Usage of Diffusion Models for Thermal Image Super-Resolution: A Generic, Uncertainty-Aware Approach for Guided and Non-Guided Schemes
Abstract
In this paper, we explore the use of diffusion models for the thermal imaging super-resolution problem, with the PBVS workshop Thermal Image Super-Resolution Challenge (TISR) as an application context. In addition of adapting the recently proposed Resshift diffusion approach to the problem of SR for thermal imaging, we show how this diffusion model can be also used nearly effortless in both the guided and non-guided TISR tasks, where the guidance comes from the visible image. More crucially, we show that a natural and often under-leveraged output from this diffusion approach is the quantification of the aleatoric uncertainty on the resulting HR prediction. By using this property, we empirically show that per-pixel standard deviation of the samples produced by a super-resolution diffusion model are a good estimator for the per-pixel absolute error in scenarios where the HR ground truth is not available.
Cite
Text
Cortés-Mendez and Hayet. "Exploring the Usage of Diffusion Models for Thermal Image Super-Resolution: A Generic, Uncertainty-Aware Approach for Guided and Non-Guided Schemes." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2024. doi:10.1109/CVPRW63382.2024.00318Markdown
[Cortés-Mendez and Hayet. "Exploring the Usage of Diffusion Models for Thermal Image Super-Resolution: A Generic, Uncertainty-Aware Approach for Guided and Non-Guided Schemes." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2024.](https://mlanthology.org/cvprw/2024/cortesmendez2024cvprw-exploring/) doi:10.1109/CVPRW63382.2024.00318BibTeX
@inproceedings{cortesmendez2024cvprw-exploring,
title = {{Exploring the Usage of Diffusion Models for Thermal Image Super-Resolution: A Generic, Uncertainty-Aware Approach for Guided and Non-Guided Schemes}},
author = {Cortés-Mendez, Carlos and Hayet, Jean-Bernard},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2024},
pages = {3123-3130},
doi = {10.1109/CVPRW63382.2024.00318},
url = {https://mlanthology.org/cvprw/2024/cortesmendez2024cvprw-exploring/}
}