# MBLLEN

**Repository Path**: xxxxcp/MBLLEN

## Basic Information

- **Project Name**: MBLLEN
- **Description**: Code for “MBLLEN: Low-light Image/Video Enhancement Using CNNs”, BMVC 2018.
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-11-09
- **Last Updated**: 2021-11-09

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# MBLLEN

This is a Keras implementation of “MBLLEN: Low-light Image/Video Enhancement Using CNNs” in BMVC 2018, by [Feifan Lv](https://lvfeifan.github.io/), [Feng Lu](http://shi.buaa.edu.cn/lufeng/), Jianhua Wu and Chongsoon Lim. This page provides more experiments on real low-light images compared with latest methods.

**[Paper](http://bmvc2018.org/contents/papers/0700.pdf) and [Project page](http://phi-ai.org/project/MBLLEN/default.htm)**

![](figs/figure.png)

## Requirements ##

- [x] python 3  
- [x] Tensorflow 1.6.0
- [x] Keras 2.2.0
- [x] Opencv-python 3.4.2

## Usage ##

#### Testing

To quickly test your own low-light images with our model, you can just run through

```shell
cd main
python test.py -i <input folder> -r <output folder> -m <model name>
```

By default, the code takes the data in the "../input/" folder, loads the "Syn_img_lowlight_withnoise.h5" model and saves results in the "../result/" folder.  Please read the code to see other parameter settings. 

#### Training:

First, prepare your own dataset or download our synthetic low-light dataset from our [Project page](http://phi-ai.org/project/MBLLEN/default.htm). Second, change the load images path of "train.py" and "data_load.py". Then, you can just run through

```shell
cd main
python train.py
```

By default, the code takes the data in the "../dataset/" folder and save weights in the "./models/" folder. Please read the code to see other parameter settings.


## Model

- [x] Syn_img_lowlight_withnoise.h5  (This model is trained using synthetic lowlight images based on Poisson noise model. It is using for enhancement and denoising simultaneously.)
- [x] Syn_img_lowlight.h5  (This model is trained using synthetic lowlight images without additional noise. It is using only for enhancement.)
- [x] LOL_img_lowlight.h5  (This model is fine-tuned using LOL Dataset.)

## Performance on Real Lowlight Images

To obtain better enhancement result, we linearly amplify the output of the network to improve contrast. Please read the code to see other parameter settings. 

![](figs/figure2.png)

![](figs/reallowlight.png)

##  Performance on [LOL Dataset](https://github.com/weichen582/RetinexNet)

Our LOL fine-tuned version performs well on LOL test images. 

![](figs/LOL.png)

##  Performance on [DeepUPE Dataset](https://github.com/wangruixing/DeepUPE)

Our model is comparable with DeepUPE. Notice that, our models are not fine-tuned using DeepUPE's images (training images are not provided).

![](figs/Deep.png )

## Video Version

Coming Soon ...

## Bibtex

If you use this code for your research, please cite our paper.

 ```
 @inproceedings{Lv2018MBLLEN,
  title={MBLLEN: Low-light Image/Video Enhancement Using CNNs},
  author={Feifan Lv, Feng Lu, Jianhua Wu, Chongsoon Lim},
  booktitle={British Machine Vision Conference (BMVC)},
  year={2018}
}
 ```

## Related Follow-Up Work

Feifan Lv, Yu Li and Feng Lu. Attention-guided Low-light Image Enhancement. arXiv:1908.00682, 2019. **[Paper](https://arxiv.org/abs/1908.00682) and [Project page](http://phi-ai.org/project/AgLLNet/default.htm)**