dabria531 / rainremoval_iccv2017

Should We Encode Rain as Deterministic or Stochastic?

Rain Removal Code

Version 1.0, 17-Aug-2017

This package contains the MATLAB implementation of "Should We Encode Rain Streaks in Video as Deterministic or Stochastic?".

The code has been tested with MATLAB 2016b on a Mac.

Use of this code is free for research purposes only. Shall not be used for commercial purposes! All copyrights belong to the original authors. The technology has applied for patents. If you want to purchase the patents for commercial purposes, please contact the corresponding author: Deyu Meng ([email protected]). Thank you!

Reference:

Please cite this paper if you use the code:

Wei Wei, Lixuan Yi, Qi Xie, Qian Zhao, Deyu Meng, Zongben Xu, Should We Encode Rain Streaks in Video as Deterministic or Stochastic? 16th International Conference on Computer Vision(ICCV), Venice, Italy, Oct. 2017.

Usage:

1. Unpack the contents of the compressed file to a new directory.

2. Run the DEMO.

Dataset:

Synthetic rain video: ‘girl.mat’, ‘pedestrian.mat’, ‘truck.mat’, ‘highway.mat’, corresponding to Fig. 4-7 in the reference.

Real rain video: ‘wall.mat’, ‘traffic.mat’, corresponding to Fig. 8-9 in the reference, with and without moving objects, respectively.

Acknowledgments

Code partly borrows from:

[1] Deyu Meng, Fernando De la Torre. Robust Matrix Factorization with Unknown Noise. ICCV, 2013.

[2] Xiaowei Zhou, Can Yang, Weichuan Yu. Moving Object Detection by Detecting Contiguous Outliers in the Low-Rank Representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(3):597–610, 2013.

Thanks for sharing!

If you have any questions, please feel free to contact Wei Wei ([email protected]).