Fusion of algorithms for Compressed Sensing

Ambat, Sooraj K.; Chatterjee, Saikat; Hari, K.V.S.

doi:10.1109/icassp.2013.6638788

Cited by 16 publications

(16 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, N }. For the measurement of ASRER, Gaussian noise with SMNR = 41 dB was added to the measurements, where SMNR (signal to measurement noise ratio) is defined as in [7], SM N R =…”

Section: Resultsmentioning

confidence: 99%

“…Some of the most popular algorithms for the reconstruction problem are, Basis Pursuit (a convex relaxation to l 0 norm minimization) [1], Matching Pursuit (MP) [2], Orthogonal Matching Pursuit (OMP) [3], Look Ahead OMP (LAOMP) [4], Subspace Pursuit (SP) [5], CoSaMP [6], FACS [7] etc. These algorithms can be broadly classified into Convex relaxation methods [1] and greedy pursuit algorithms [2]- [7]. In convex relaxation methods, the 1 norm of the vector to be reconstructed is minimized, subject to the constraint that the reconstructed vector gives the same measurement vector under the sampling.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Batch Look Ahead Orthogonal Matching Pursuit

Muralikrishnna

Ambat

Hari

2018

2018 Twenty Fourth National Conference on Communications (NCC)

Self Cite

View full text Add to dashboard Cite

Compressed sensing (CS) is a sampling paradigm that enables sampling signals at sub Nyquist rates by exploiting the sparse nature of signals. One of the main concerns in CS is the reconstruction of the signal after sampling. Many reconstruction algorithms have been proposed in the literature for the recovery of the sparse signals-Basis Pursuit, Orthogonal Matching Pursuit (OMP), Look Ahead Orthogonal Matching Pursuit (LAOMP) are some of the popular reconstruction algorithms. LAOMP, a modification of OMP, improves the reconstruction accuracy of OMP by employing a look ahead procedure. But LAOMP suffers from the drawback of being very expensive in terms of the computational time. In this paper we propose a modified version of the LAOMP algorithm called Batch-LAOMP which has a lesser computational complexity and also gives better performance in terms of reconstruction accuracy as seen from the results of the numerical experiments.

show abstract

“…, N }. For the measurement of ASRER, Gaussian noise with SMNR = 41 dB was added to the measurements, where SMNR (signal to measurement noise ratio) is defined as in [7], SM N R =…”

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Batch Look Ahead Orthogonal Matching Pursuit

Muralikrishnna

Ambat

Hari

2018

2018 Twenty Fourth National Conference on Communications (NCC)

Self Cite

View full text Add to dashboard Cite

show abstract

“…1 stands for the vector 1-norm. The GPABP algorithm was shown to outperform fusion-based CS reconstruction algorithms such as fusion-of-algorithms for CS [21] and committee machine approach for CS [22].…”

Section: Prior Workmentioning

confidence: 99%

Greedy Pursuits Based Gradual Weighting Strategy for Weighted <tex>$\ell_{1}$</tex>-Minimization

Narayanan

Sahoo

Makur

2018

2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

In Compressive Sensing (CS) of sparse signals, standard 1minimization can be effectively replaced with Weighted 1 -minimization (W 1 ) if some information about the signal or its sparsity pattern is available. If no such information is available, Re-Weighted 1 -minimization (ReW 1 ) can be deployed. ReW 1 solves a series of W 1 problems, and therefore, its computational complexity is high. An alternative to ReW 1 is the Greedy Pursuits Assisted Basis Pursuit (GPABP) which employs multiple Greedy Pursuits (GPs) to obtain signal information which in turn is used to run W 1 . Although GPABP is an effective fusion technique, it adapts a binary weighting strategy for running W 1 , which is very restrictive. In this article, we propose a gradual weighting strategy for W 1 , which handles the signal estimates resulting from multiple GPs more effectively compared to the binary weighting strategy of GPABP. The resulting algorithm is termed as Greedy Pursuits assisted Weighted 1 -minimization (GP-W 1 ). For GP-W 1 , we derive the theoretical upper bound on its reconstruction error. Through simulation results, we show that the proposed GP-W 1 outperforms ReW 1 and the state-of-the-art GPABP.

show abstract

“…Without prior knowledge, one cannot judiciously choose the best sparse reconstruction algorithm from a set of viable algorithms. FACS proposed in [4] fuses the estimates from several sparse reconstruction algorithms to form a better estimate.…”

Section: Background : Cs and Facsmentioning

confidence: 99%

“…To address these issues a fusion framework named 'Fusion of Algorithms for Compressed Sensing (FACS)' was proposed in [4] . But it has been observed that in low dimension measurement regime 'FACS' ends up solving an ill-conditioned least squares problem leading to poor reconstruction capability.…”

Section: Introductionmentioning

confidence: 99%

Modified greedy pursuits for improving sparse recovery

Deepa

Ambat

Hari

2014

2014 Twentieth National Conference on Communications (NCC)

Self Cite

View full text Add to dashboard Cite

Compressive Sensing (CS) theory combines the signal sampling and compression for sparse signals resulting in reduction in sampling rate. In recent years, many recovery algorithms have been proposed to reconstruct the signal efficiently. Subspace Pursuit and Compressive Sampling Matching Pursuit are some of the popular greedy methods. Also, Fusion of Algorithms for Compressed Sensing is a recently proposed method where several CS reconstruction algorithms participate and the final estimate of the underlying sparse signal is determined by fusing the estimates obtained from the participating algorithms. All these methods involve solving a least squares problem which may be ill-conditioned, especially in the low dimension measurement regime. In this paper, we propose a step prior to least squares to ensure the well-conditioning of the least squares problem. Using Monte Carlo simulations, we show that in low dimension measurement scenario, this modification improves the reconstruction capability of the algorithm in clean as well as noisy measurement cases.

show abstract

Fusion of algorithms for Compressed Sensing

Cited by 16 publications

References 16 publications

Batch Look Ahead Orthogonal Matching Pursuit

Batch Look Ahead Orthogonal Matching Pursuit

Greedy Pursuits Based Gradual Weighting Strategy for Weighted <tex>$\ell_{1}$</tex>-Minimization

Modified greedy pursuits for improving sparse recovery

Contact Info

Product

Resources

About