This is a MATLAB implementation based of the CoLaMP algorithms presented in the following paper (download). "Estimating Sparse Signals with Smooth Support via Convex Programming and Block Sparsity", Sohil Atul Shah, Christoph Studer and Tom Goldstein
The source code and datasets are published under the MIT Licence. See LICENSE for details. In general you can use them for any purpose with proper attribution. If you do something interesting with the code, we'll be happy to know about it. Feel free to contact us.
As presented in the paper Algorithm 1 is implemented as part of block-sparse-RPCA code and Algorithm 2 is implemented in matching_pursuit.m. Apart from this, we also provide ADMM implementation for Algorithm 1 here. The denoising algorithm is implemented by plugging in Primal-Dual algorithm of this paper.
We have included test wrapper for all the four application in the codebase for the ease of understanding the input/output and for reproducing some of the paper's qualitative results.