Compressive MUSIC, Forward/Backward Compressive Subspace Fitting Implementations

Jong Chul Ye alerted me to this new resource:

".....Hi Igor,

....  I would like to bring your attention to our website with a compressive sensing joint sparse recovery software packages for multiple measurement vector (MMV) problems.

      http://bisp.kaist.ac.kr/joint_sparse_recovery_v2.html

The current version of package includes three algorithms to address MMV problems:

1. Compressive MUSIC (CS-MUSIC)
2. Forward Compressive Subspace Fitting (forward CSF)
3. Backward Compressive Subspace Fitting (backward CSF)

More specifically, Compressive MUSIC identifies the parts of support using CS, after which the remaining supports are estimated using a novel generalized MUSIC criterion. Using a large system MMV model, we showed that our compressive MUSIC requires a smaller number of sensor elements for accurate support recovery than the existing CS methods and that it can approach the optimal L0-bound with finite number of snapshots. The theoretical analysis of Compressive MUSIC will appear in 2012 January Issue of IEEE Trans. on Information Theory.

While these type of hybrid algorithms are optimal for critically sampled cases, they have limitations in exploiting the redundant sampling to improve noise robustness. To address this issue,   we recently introduced a novel subspace fitting criterion that extends the generalized MUSIC criterion so that it exhibits near-optimal behaviors for various sampling conditions. In addition, the subspace fitting criterion leads to two alternative forms of compressive subspace fitting (CSF) algorithms with forward and backward support selection (forward CSF, and backward CSF), which significantly improve the noise robustness.

Related manuscripts can be downloaded from the following links:

[1] J. M. Kim, O. K. Lee, and J. C. Ye. Compressive MUSIC: A Missing Link between Compressive Sensing and Array Signal Processing. IEEE Trans. Inf. Theory, Jan. 2012 [preprint is here]

[2] J. M. Kim, O. K. Lee, and J. C. Ye, 2011. Noise Robust Joint Sparse Recovery using Compressive Subspace Fitting.  arXiv:1112.3446v1 [cs.IT]

If you have any question, please feel free to contact me.

Happy New Year !

-Jong..."

Thanks  Jong . This webpage will be both featured in the Big Picture in Compressive Sensing and in the Matrix Factorization Jungle Page..The earlier mention "Compressive MUSIC with optimized partial support for joint sparse recovery by Jong Min Kim, Ok Kyun Lee, Jong Chul Ye [no code]" will now be replaced with a link to the page where the code resides.

Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.