shizenglin / Unsharp-Mask-Guided-Filtering

Code for "Unsharp Mask Guided Filtering, IEEE Transactions on Image Processing, 2021"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

image

image

Unsharp-mask guided filtering without learning


You can find the code in the folder of "matlab"

Usage


Run "example_smoothing.m", "example_enhancement.m", and "example_flash.m" to get the Figures 3, 4 and 5

Unsharp-mask guided filtering with learning


You can find the code in the folder of "learning"

Requirements

1. CUDA 8.0 and Cudnn 7.5 or higher
2. GPU memory 10GB or higher
3. Python 2.7 or higher
4. Tensorflow 2.0 or higher. If your Tensorflow version is lower than 2.0, you should replace "import tensorflow.compat.v1 as tf" with "import tensorflow as tf" in "main.py", "model.py", and "ops.py"

Training

1. Prepare your data (download the NYU Depth V2 dataset here) following Section V-A.
2. Set the experiment settings in ¨tr_param.ini¨ in which phase = train, and set other parameters accordingly (refer to our paper).
3. Run ¨python main.py¨

Testing

1. Prepare your data following Section V-A.
2. Set the experiment settings in ¨tr_param.ini¨ in which phase = test, and set other parameters accordingly (refer to our paper).
3. Run ¨python main.py¨

Please cite our paper when you use this code.

 @article{shi2021unsharp,
    title={Unsharp Mask Guided Filtering},
    author={Shi, Zenglin and Chen, Yunlu and Gavves, Efstratios and Mettes, Pascal and Snoek, Cees GM},
    journal={IEEE Transactions on Image Processing},
    volume={30},
    pages={7472 - 7485},
    year={2021},
    publisher={IEEE}
 }

About

Code for "Unsharp Mask Guided Filtering, IEEE Transactions on Image Processing, 2021"


Languages

Language:Python 91.2%Language:MATLAB 8.8%