Matlab Software for Recent Research (Modified-CS, LS-CS and KF-CS, ReProCS)

 

Modified-CS Code (written by Wei Lu, greatjackylu@gmail.com )

Please cite following papers when you use this code

o   Namrata Vaswani and Wei Lu, Modified-CS: Modifying Compressive Sensing for Problems with Partially Known Support,   IEEE Trans. Signal Processing, Vol. 58, No. 9, September, 2010. Shorter version in ISIT 2009.

o   Wei Lu, Taoran Li, Ian Atkinson,  Namrata Vaswani Modified-CS-Residual for Recursive Reconstruction of Highly Undersampled Functional MRI Sequences, IEEE Intl. Conf. Image Proc. (ICIP) 2011

o   Wei Lu and Namrata Vaswani, Regularized Modified BPDN for Noisy Sparse Reconstruction with Partial Erroneous Support and Signal Value Knowledge, IEEE Trans. Signal Processing, Vol. 60, No. 1, January, 2012

o   Wei Lu and Namrata Vaswani, Exact Reconstruction Conditions for Regularized Modified Basis Pursuit, IEEE Trans. Signal Processing, vol. 60, No. 5, May, 2012


Modified-CS(noiseless): CVX based code: small sized signal/image sequence (smaller than 64X64 images)

https://shared.com/j7p195rfsa?s=l

A note about ModifiedCS_sequential.m

- The following may appear as a small typo: for seq=1:seqlen should be replaced by for seq=2:seqlen

- But it does not affect results in any way for sparse sequences, it actually improves results for compressible sequences.

 

Modified-CS(noiseless): For large sized signal/image sequence: Optimization algorithm coded in directly using fast operators for 2D-DFT and 2D-DWT - does not need to store the measurement matrices ever (of course also works for small sized signals/images).
https://shared.com/bhl97t3raf?s=l

 


Modified-CS-residual(noisy):

Optimization algorithm coded in directly - same idea as above, works also for large sized signals/images.
https://shared.com/ok111wu3dg?s=l

 


Modified-CS-residual for fMRI:

Optimization algorithm coded in directly - same idea are above, works also for large sized/ signals/images.
https://shared.com/7t1yqj9g9m?s=l

 

 

Kalman filtered Compressed Sensing (KF-CS) and Least Squares CS residual(LS-CS) code

Please cite following papers when you use this code

o   Namrata Vaswani, LS-CS-residual (LS-CS): Compressive Sensing on the Least Squares Residual, IEEE Trans. Signal Processing, Vol. 58, No. 8, August, 2010.

o   Namrata Vaswani, Kalman Filtered Compressed Sensing, IEEE Intl. Conf. Image Proc. (ICIP), 2008.

o   Namrata Vaswani, Analyzing Least Squares and Kalman Filtered Compressed Sensing, IEEE Intl. Conf. Acous. Speech. Sig. Proc. (ICASSP), 2009.

 

Latest code for KF-CS and LS-CS: Code ReadMe file

Contains LS-CS and two versions of KF-CS: KF-CS-LS and KF-CS-KF, both with and without deletion steps.

 

Two older versions of the code

Kalman filtered CS (KF-CS): KFCS_new.zip (Main file: runsims_final, see comments and see README.txt)

See README.txt for code structure. runsims_final.m is the main file. kfcs_full contains the kfcs code.

Least Squares CS (LS-CS): Replace the KF in the above code by LS: to get the LS-CS implementation

 

Oldest version of KF-CS code based on the ICIP'08 paper: KFCS.zip (To run it: runsims2, followed by plotting the errors)

See README.txt or comments in runsims2.m

You may need to install netlab and add it to your MATLAB path: netlab.zip

 

 

Recursive Projected CS or ReProCS for Recursive Robust PCA or Recursive Sparse Recovery in Large but Structured Noise

Please cite the following papers when you use this code

o   Chenlu Qiu and Namrata Vaswani, Real-time Robust Principal Components' Pursuit, Allerton, 2010

o   Chenlu Qiu and Namrata Vaswani, Recursive Sparse Recovery in Large but Correlated Noise, Allerton 2011

o   Chenlu Qiu, Namrata Vaswani, Brian Lois and Leslie Hogben, Recursive Robust PCA or Recursive Sparse Recovery in Large but Structured Noise, IEEE Trans. Information Theory, August 2014

o   Han Guo, Chenlu Qiu and Namrata Vaswani, An Online Algorithm for Separating Sparse and Low-dimensional Signal Sequences from their Sum, IEEE Trans. Signal Processing, August, 2014

 

Latest Code link for practical ReProCS: http://www.ece.iastate.edu/~hanguo/PracReProCS.html#Code_, code

 Older code: http://www.ece.iastate.edu/~chenlu/ReProCS/ReProCS_main.htm#Matlab_code

 

Older work

o        Code for my older work is obtained by following this link: http://www.ece.iastate.edu/~namrata/research/research.html