https://github.com/vinthony/depth-distillation
[ECCV 2020] Defocus Blur Detection via Depth Distillation
https://github.com/vinthony/depth-distillation
defocus-blur defocus-blur-detection defocus-estimation
Last synced: about 1 year ago
JSON representation
[ECCV 2020] Defocus Blur Detection via Depth Distillation
- Host: GitHub
- URL: https://github.com/vinthony/depth-distillation
- Owner: vinthony
- License: mit
- Created: 2020-06-01T02:12:16.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2022-03-18T12:41:34.000Z (about 4 years ago)
- Last Synced: 2025-04-03T08:02:49.501Z (about 1 year ago)
- Topics: defocus-blur, defocus-blur-detection, defocus-estimation
- Language: Python
- Homepage: https://arxiv.org/pdf/2007.08113.pdf
- Size: 37.1 KB
- Stars: 70
- Watchers: 8
- Forks: 14
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Defocus Blur Detection via Depth Distillation
This repo contains the code and results of our ECCV 2020 paper:
[Defocus Blur Detection via Depth Distillation](https://arxiv.org/abs/2007.08113)
[_Xiaodong Cun_](http://vinthony.github.io) and [_Chi-Man Pun_*](http://www.cis.umac.mo/~cmpun/)
[University of Macau](http://um.edu.mo/)
[Models](#pretrained-models) | [Results](#results) | [Paper](https://arxiv.org/abs/2007.08113) | [Supp.](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123580732-supp.pdf) | [Online Demo!(Google CoLab)](https://colab.research.google.com/drive/1a-Un_lZqkEN-mr-SzQh9GLy4qXIJgn0v#scrollTo=Lh2_NGuLaM_c)
### Results
we provide results on two datasets under different backbone(VGG19,ResNext101), please download from [Google Drive](https://drive.google.com/file/d/13nhzz9qaQ6U0F7Jsu4mLMS8XktZK1-Y_/view?usp=sharing)
### Pretrained Models
* VGG16 backbone: [vgg_best.pth](https://drive.google.com/file/d/1VigqrPdiIF18VALo92L9WCuASnpzu7qa/view?usp=sharing)
* ResNeXt101 backbone: [res_best.pth](https://drive.google.com/file/d/1HrzFqXSJY2Zxd9qvdKC7_Glxljjd27sf/view?usp=sharing)
### Dependences
* PyTorch
* OpenCV
* scipy
* tqdm
* scikit-learn
### Demos
[Online Demo!(Google CoLab)](https://colab.research.google.com/drive/1a-Un_lZqkEN-mr-SzQh9GLy4qXIJgn0v#scrollTo=Lh2_NGuLaM_c) is recommanded to evaluate the performance of our method.
Also, you can run a local jupyter server to evalute on CPU or GPU.
1. Download the [pretrianed models](#pretrained-models) and [ResNeXt101 backbone](https://drive.google.com/file/d/1o7zQvDef9mAEzbQeHAwMSi9nK_QEhhVZ/view?usp=sharing) and put it to `pretrained`.
2. Download the [DUT500 dataset](https://drive.google.com/file/d/1Qmcu6GDgKhiHVgLxeQg23tfy5I1Xg5Xk/view?usp=sharing) and put it to `dataset`
3. make sure all the path in `paths.py` are correct, the folder may like:
```
depth-distillation/
- datasets/
* DUTDefocus/
* CUHKDefocus/
- pretrained/
* res_best.pth
* vgg_best.pth
* resnext_101_32x4d.pth
- models/
other files...
```
4. run the jupyter notebook to evaluate.
## **Acknowledgements**
The author would like to thanks Nan Chen for her helpful discussion.
Part of the code is based upon [Pytorch-GAN](https://github.com/eriklindernoren/PyTorch-GAN) and [Shadow Detection](https://github.com/zijundeng/BDRAR)
## **Citation**
If you find our work useful in your research, please consider citing:
```
@misc{cun2020defocus,
title={Defocus Blur Detection via Depth Distillation},
author={Xiaodong Cun and Chi-Man Pun},
year={2020},
eprint={2007.08113},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## **Contact**
Please contact me if there is any question (Xiaodong Cun yb87432@um.edu.mo)