hanxuhfut / CVPR2019_Underexposed_Photo_Enhancement_Using_Deep_Illumination_Estimation

Underexposed Photo Enhancement Using Deep Illumination Estimation

Ruixing Wang¹, Qing Zhang², Chi-Wing Fu¹, Xiaoyong Shen³, Wei-Shi Zheng², Jiaya Jia^1,3

¹The chinese university of hong kong ²Sun Yat-sen University ³Tencent Youtu Lab

Thanks to the original author for sharing the code. This is my own modified version. Please refer to the author's homepage for details.

I modified the main/Makefile so that the compilation is successful. The modified "Makefile" is already in ‘main’ and I have compiled it in ‘build’.

Tested on Python 2.7, Ubuntu 14.0, gcc-4.8, CUDA 8.0, CUDNN 5.1, GeForce GTX TITAN X.

1) cd main
2) pip install -r requirements.txt
3) make

3. Evaluation: The test set can be downloaded in https://drive.google.com/file/d/1HZnNgptNxjKJAhekz2K5yh0mW0yKIws2/view?usp=sharing. It includes 500 pair images from MIT-Adobe FiveK 4500-5000. You can download this and run:

    python main/run.py checkpoints <input file path> <output file path>

PSNR evaluation code is in avg_psnr.m. Modify the related paths in 'avg_psnr.m', and run it.

Bibtex

@InProceedings{Wang_2019_CVPR,
author = {Wang, Ruixing and Zhang, Qing and Fu, Chi-Wing and Shen, Xiaoyong and Zheng, Wei-Shi and Jia, Jiaya},
title = {Underexposed Photo Enhancement Using Deep Illumination Estimation},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}
}