Shuwei Yue and *Minchen Wei
Color, Imaging, and Metaverse Research Center, The Hong Kong Polytechnic University.
Abstract:
Conventional illuminant estimation methods were developed for scenes with a uniform illumination, while recently developed methods, such as pixel-wise methods, estimate the illuminants at the pixel level, making them applicable to a wider range of scenes. It was found that the same pixel-wise algorithm had very different performance when applied to images with different bit-depths, with up to a 30% decrease in accuracy for images having a lower bit-depth. Image signal processing (ISP) pipelines, however, prefer to deal with images with a lower bit-depth. In this paper, the analyses show that such a reduction was due to the loss of details and increase of noises, which were never identified in the past. We propose a method combining the L1 loss optimization and physical-constrained post-processing. The proposed method was found to result in around 40% higher estimation accuracy, in comparison to the state-of-the-art DNN-based methods.
Main results:
Overview:
If you use this code, please cite our paper:
@article{Yue24robust,
title = {Robust pixel-wise illuminant estimation algorithm for images with a low bit-depth},
author = {Shuwei Yue and Minchen Wei},
journal = {Opt. Express},
number = {15},
pages = {26708--26718},
volume = {32},
month = {Jul},
year = {2024},
publisher = {Optica Publishing Group}
-
Download the LSMI dataset(The origin dataset is large which may taken you more than one day to process, I released the processed tiff.(~30GB) Then, put it into fold 'LSMI_data'.
-
Download the pre-trained models(~1GB) and put them into the 'pretrained_models' fold
The Net architecture is the same as LIMIU
Our key contribution is using L1 loss for fine-tuning when training and the post_processing.py when testing, as the physical-constrained post-processing, detailed in the paper.
Check the path in setting.py and run train.py
Check the path in test.py and run test.py. Default is using post-processing in USING_POST_PROCESSING=True, you can change it to False for comparison, and you will see an amazing improvement!
*Don't hesistate submit an issue if you have any questions!