Home

Awesome

Denoising Diffusion Post-Processing for Low-Light Image Enhancement


News


paper License: CC BY-NC-SA 4.0

Abstract: Low-light image enhancement (LLIE) techniques attempt to increase the visibility of images captured in low-light scenarios. However, as a result of enhancement, a variety of image degradations such as noise and color bias are revealed. Furthermore, each particular LLIE approach may introduce a different form of flaw within its enhanced results. To combat these image degradations, post-processing denoisers have widely been used, which often yield oversmoothed results lacking detail. We propose using a diffusion model as a post-processing approach, and we introduce Low-light Post-processing Diffusion Model (LPDM) in order to model the conditional distribution between under-exposed and normally-exposed images. We apply LPDM in a manner which avoids the computationally expensive generative reverse process of typical diffusion models, and post-process images in one pass through LPDM. Extensive experiments demonstrate that our approach outperforms competing post-processing denoisers by increasing the perceptual quality of enhanced low-light images on a variety of challenging low-light datasets. Source code is available at https://github.com/savvaki/LPDM.

Example results Example results Example results


Download Results

The image results for all techniques and datasets are available for download here.


Requirements

Set up and activate the virtual conda environment:


Test

Calculate Metrics


Train


BibTeX Citation

@article{panagiotou2024denoising,
title = {Denoising diffusion post-processing for low-light image enhancement},
journal = {Pattern Recognition},
volume = {156},
pages = {110799},
year = {2024},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2024.110799},
url = {https://www.sciencedirect.com/science/article/pii/S0031320324005508},
author = {Savvas Panagiotou and Anna S. Bosman}
}

References

This repository is a derivative of the original Stable Diffusion repository.