From Fidelity to Perceptual Quality: A Semi-Supervised Approach for Low-Light Image Enhancement (CVPR'2020)
Wenhan Yang, Shiqi Wang, Yuming Fang, Yue Wang and Jiaying Liu
[Paper Link] [Project Page] [Slides](TBA)[Video](TBA) (CVPR'2020 Poster)
Under-exposure introduces a series of visual degradation, i.e. decreased visibility, intensive noise, and biased color, etc. To address these problems, we propose a novel semi-supervised learning approach for low-light image enhancement. A deep recursive band network (DRBN) is proposed to recover a linear band representation of an enhanced normal-light image with paired low/normal-light images, and then obtain an improved one by recomposing the given bands via another learnable linear transformation based on a perceptual quality-driven adversarial learning with unpaired data. The architecture is powerful and flexible to have the merit of training with both paired and unpaired data. On one hand, the proposed network is well designed to extract a series of coarse-to-fine band representations, whose estimations are mutually beneficial in a recursive process. On the other hand, the extracted band representation of the enhanced image in the first stage of DRBN (recursive band learning) bridges the gap between the restoration knowledge of paired data and the perceptual quality preference to real high-quality images. Its second stage (band recomposition) learns to recompose the band representation towards fitting perceptual properties of highquality images via adversarial learning. With the help of this two-stage design, our approach generates the enhanced results with well reconstructed details and visually promising contrast and color distributions. Extensive evaluations demonstrate the superiority of our DRBN.
@InProceedings{Yang_2020_CVPR,
author = {Yang, Wenhan and Wang, Shiqi and Fang, Yuming and Wang, Yue and Liu, Jiaying},
title = {From Fidelity to Perceptual Quality: A Semi-Supervised Approach for Low-Light Image Enhancement},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}
- Clone this repo
- Install PyTorch and dependencies from http://pytorch.org
- For stage II training, you need to download https://download.pytorch.org/models/vgg16-397923af.pth and put it in DRBL-stage2/src/.
- For testing, you can directly run test.sh in DRBL-stage1/src/ and DRBL-stage2/src/.
- For training, you can directly run train.sh in DRBL-stage1/src/ and DRBL-stage2/src/.
- You can download our dataset here: https://pan.baidu.com/s/1-j6_3G9WHS8rkEBzHN7QvQ (extracted code: r0xd)
（Note: the code is suitable for PyTorch 0.4.1）
If you have questions, you can contact yangwenhan@pku.edu.cn.