A Trilateral Weighted Sparse Coding Scheme for Real-World Image Denoising, ECCV 2018.
Switch branches/tags
Nothing to show
Clone or download
Latest commit 899807d Oct 20, 2018
Failed to load latest commit information.
WSCandSC_notoptimized add Jul 13, 2018
.DS_Store ad d Aug 13, 2018
Block_Matching.m add Jul 13, 2018
Block_Matching_RW.m add Jul 13, 2018
ChangeLog.txt add Jul 13, 2018
Demo_TWSC_RGB2YUV_CC2016.m add Jul 13, 2018
Demo_TWSC_RGB2YUV_DND2017.m add Jul 13, 2018
Demo_TWSC_Sigma_AWGN.m add Jul 13, 2018
Demo_TWSC_Sigma_RW_CC2016.m add Jul 13, 2018
Demo_TWSC_Sigma_RW_DND2017.m add Jul 13, 2018
Demo_TWSC_Sigma_RW_NC.m add Jul 13, 2018
Image2Patch.m add Jul 13, 2018
License.txt add Jul 13, 2018
NoiseEstimation.m add Jul 13, 2018
PGs2Image.m add Jul 13, 2018
Poster_TWSC_ECCV2018.pdf add Poster Oct 20, 2018
README.md add Jul 13, 2018
SearchNeighborIndex.m add Jul 13, 2018
TWSC_ADMM.m add Jul 13, 2018
TWSC_Sigma_AWGN.m add Oct 10, 2018
TWSC_Sigma_RW.m add Oct 10, 2018
cal_ssim.m add Jul 13, 2018
csnr.m add Jul 13, 2018
image2cols.m add Jul 13, 2018


% ===============================================================

The code in this package implements the Trilateral Weighted Sparse Coding Scheme for real color image denoising as described in the following paper:


     author = {Jun Xu and Lei Zhang and David Zhang}, 

     title = {A Trilateral Weighted Sparse Coding Scheme for Real-World Image Denoising}, 

     journal = {ECCV},

     year = {2018} 


Please cite the paper if you feel this code useful in your research. Please see the file License.txt for the license governing this code.

Version: 1.0 (13/07/2018), see ChangeLog.txt Contact: Jun Xu <csjunxu@comp.polyu.edu.hk, nankaimathxujun@gmail.com>

% ===============================================================


Code will be updated on https://github.com/csjunxu/TWSC-ECCV2018


Please download the data from corresponding addresses.

  1. cleanimages: 20 high quality commonly used natural gray scale images

  2. nc: real noisy images with no ''ground truth'' This dataset can be found at http://demo.ipol.im/demo/125/

  3. cc: 15 cropped real noisy images from CC [1]. This dataset can be found at http://snam.ml/research/ccnoise The smaller 15 cropped images can be found on in the directory ''Real_ccnoise_denoised_part'' of https://github.com/csjunxu/MCWNNM_ICCV2017 The *real.png are noisy images; The *mean.png are "ground truth" images; The *ours.png are images denoised by CC.

  4. dnd: The Darmstadt Noise Dataset [2] consists of 50 pairs of real noisy images, each images provides 50 crops, resulting overall 1,000 crops provided on https://noise.visinf.tu-darmstadt.de/

[1] A Holistic Approach to Cross-Channel Image Noise Modeling and its Application to Image Denoising. Seonghyeon Nam*, Youngbae Hwang*, Yasuyuki Matsushita, Seon Joo Kim. CVPR 2016.

[2] Benchmarking Denoising Algorithms with Real Photographs. Tobias Plötz and Stefan Roth. CVPR 2017.


This code is implemented purely in Matlab2014b and doesn't depends on any other toolbox.


If you have any questions or suggestions with the code, or find a bug, please let us know. Contact Jun Xu at csjunxu@comp.polyu.edu.hk or nankaimathxujun@gmail.com.