Trust-region methods for nonconvex sparse recovery optimization

We solve the ℓ2-ℓp sparse recovery problem by transforming the objective function into an unconstrained differentiable function and apply a limited-memory trust-region method. Unlike gradient projection-type methods, which uses only the current gradient, our approach uses gradients from previous iterations to obtain a more accurate Hessian approximation. Preliminary numerical experiments with simulated compressive sensing 1D data are provided to illustrate that our proposed approach eliminates spurious solutions more effectively while improving the computational time to converge in comparison to standard approaches.

[1]  D K Smith,et al.  Numerical Optimization , 2001, J. Oper. Res. Soc..

[2]  Mário A. T. Figueiredo,et al.  Gradient Projection for Sparse Reconstruction: Application to Compressed Sensing and Other Inverse Problems , 2007, IEEE Journal of Selected Topics in Signal Processing.

[3]  Jingjie Cao,et al.  Recovery of seismic wavefields based on compressive sensing by an l1-norm constrained trust region method and the piecewise random subsampling , 2011, Geophysical Journal International.

[4]  S. V. N. Vishwanathan,et al.  A Quasi-Newton Approach to Nonsmooth Convex Optimization Problems in Machine Learning , 2008, J. Mach. Learn. Res..

[5]  Qiang Zhang,et al.  Randomized SVD Methods in Hyperspectral Imaging , 2012, J. Electr. Comput. Eng..

[6]  Michael Hintermüller,et al.  Nonconvex TVq-Models in Image Restoration: Analysis and a Trust-Region Regularization-Based Superlinearly Convergent Solver , 2013, SIAM J. Imaging Sci..

[7]  R. Chartrand,et al.  Restricted isometry properties and nonconvex compressive sensing , 2007 .

[8]  Mohamed-Jalal Fadili,et al.  A quasi-Newton proximal splitting method , 2012, NIPS.

[9]  Rick Chartrand,et al.  Exact Reconstruction of Sparse Signals via Nonconvex Minimization , 2007, IEEE Signal Processing Letters.

[10]  Chun Qi,et al.  Quasi-Newton Iterative Projection Algorithm for Sparse Recovery , 2014, Neurocomputing.

[11]  James E. Fowler,et al.  Compressive-Projection Principal Component Analysis , 2009, IEEE Transactions on Image Processing.

[12]  David M. author-Gay.  Computing Optimal Locally Constrained Steps , 1981 .

[13]  Rebecca Willett,et al.  Compressive coded aperture superresolution image reconstruction , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[14]  S. Prasad,et al.  Shape, Pose, and Material Recovery of Solar-Illuminated Surfaces from Compressive Spectral-Polarimetric Image Data , 2013 .

[15]  Qiang Zhang,et al.  Image Reconstruction From Double Random Projection , 2014, IEEE Transactions on Image Processing.

[16]  Jorge Nocedal,et al.  Representations of quasi-Newton matrices and their use in limited memory methods , 1994, Math. Program..

[17]  Jorge J. Moré,et al.  Computing a Trust Region Step , 1983 .

[18]  Roummel F. Marcia,et al.  Trust-Region Methods for Sparse Relaxation , 2016, 1602.08813.

[19]  Qiang Zhang,et al.  Joint segmentation and reconstruction of hyperspectral data with compressed measurements. , 2011, Applied optics.

[20]  Rebecca Willett,et al.  Regularized Non-Gaussian Image Denoising , 2015, 1508.02971.

[21]  Ashwin A. Wagadarikar,et al.  Single disperser design for coded aperture snapshot spectral imaging. , 2008, Applied optics.

[22]  Emmanuel J. Candès,et al.  Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information , 2004, IEEE Transactions on Information Theory.

[23]  Wotao Yin,et al.  Iteratively reweighted algorithms for compressive sensing , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[24]  M E Gehm,et al.  Single-shot compressive spectral imaging with a dual-disperser architecture. , 2007, Optics express.

[25]  Nicholas I. M. Gould,et al.  Trust Region Methods , 2000, MOS-SIAM Series on Optimization.

[26]  David Zhang,et al.  A Generalized Iterated Shrinkage Algorithm for Non-convex Sparse Coding , 2013, 2013 IEEE International Conference on Computer Vision.

[27]  Qiang Zhang,et al.  Randomized methods in lossless compression of hyperspectral data , 2013 .

[28]  Inderjit S. Dhillon,et al.  Clustering with Bregman Divergences , 2005, J. Mach. Learn. Res..

[29]  Sudhakar Prasad,et al.  Deblurring Compressive Spectro-Polarimetric Images Taken Trough Atmospheric Turbulence , 2014 .

[30]  Wei Dai,et al.  Low rank matrix completion: A smoothed l0-search , 2012, 2012 50th Annual Allerton Conference on Communication, Control, and Computing (Allerton).

[31]  Michael A. Saunders,et al.  Proximal Newton-type methods for convex optimization , 2012, NIPS.

[32]  Rick Chartrand,et al.  Nonconvex Compressed Sensing and Error Correction , 2007, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.