Hankel Matrix Nuclear Norm Regularized Tensor Completion for $N$-dimensional Exponential Signals

  title={Hankel Matrix Nuclear Norm Regularized Tensor Completion for \$N\$-dimensional Exponential Signals},
  author={Jiaxi Ying and Hengfa Lu and Qingtao Wei and Jian-Feng Cai and Di Guo and Jihui Wu and Zhong Chen and Xiaobo Qu},
  journal={IEEE Transactions on Signal Processing},
Signals are generally modeled as a superposition of exponential functions in spectroscopy of chemistry, biology, and medical imaging. For fast data acquisition or other inevitable reasons, however, only a small amount of samples may be acquired, and thus, how to recover the full signal becomes an active research topic, but existing approaches cannot efficiently recover <inline-formula><tex-math notation="LaTeX">$N$</tex-math></inline-formula>-dimensional exponential signals with <inline-formula… 
Vandermonde Factorization of Hankel Matrix for Complex Exponential Signal Recovery—Application in Fast NMR Spectroscopy
The Vandermonde structure of the Hankel matrix formed by the exponential signal is exploited and formulate signal recovery asHankel matrix completion with VandermondE factorization (HVaF), which is validated on biological magnetic resonance spectroscopy data.
Exponential Signal Reconstruction With Deep Hankel Matrix Factorization.
This work proposes a deep learning method whose neural network structure is designed by imitating the iterative process in the model-based state-of-the-art exponentials' reconstruction method with the low-rank Hankel matrix factorization, and demonstrates that the new method yields much lower reconstruction errors and preserves theLow-intensity signals much better than compared methods.
Complex Exponential Signal Recovery with Deep Hankel Matrix Factorization
This work proposes a deep learning method whose neural network structure is designed by unrolling the iterative process in the model-based state-of-the-art exponentials reconstruction method with low rank Hankel matrix factorization, and demonstrates that the new method yields much lower reconstruction errors and more accuracy in spectrum parameter quantification than another state- of theartDeep learning method.
A Fast Low Rank Hankel Matrix Factorization Reconstruction Method for Non-Uniformly Sampled Magnetic Resonance Spectroscopy
A low-rank matrix factorization method that avoids singular value decomposition is introduced to enable fast MRS reconstruction without sacrificing the spectra quality and enables reconstructing the challenging 3-D MRS within 15 minutes.
Improved Reconstruction of Low Intensity Magnetic Resonance Spectroscopy With Weighted Low Rank Hankel Matrix Completion
Experimental results on both synthetic and real MRS data demonstrate that the proposed approach can reconstruct low-intensity spectral peaks better than the state-of-the-art LRHM method.
An Automatic Denoising Method for NMR Spectroscopy Based on Low-Rank Hankel Model
The results on synthetic and realistic NMR spectroscopy data demonstrate a superior denoising performance of the proposed approach over typical Cadzow and the state-of-the-art QR decomposition methods, especially in the low signal-to-noise ratio regime.
Manifold Recovery Using Kernel Low-Rank Regularization: Application to Dynamic Imaging
A novel kernel low-rank algorithm to recover free-breathing and ungated dynamic MRI data from highly undersampled measurements and shows the relation of this two-step algorithm with the recent SToRM approach, thus reconciling STo RM and manifold regularization methods with algorithms that rely on explicit lifting of data to a high dimensional space.
Non-convex methods for spectrally sparse signal reconstruction via low-rank Hankel matrix completion
This work develops two fast and provable non-convex solvers, FIHT and PGD, suitable for handling spectrally sparse signals of large size, which may be prohibited for previous convex methods.
A Sparse Model-Inspired Deep Thresholding Network for Exponential Signal Reconstruction--Application in Fast Biological Spectroscopy.
  • Z. Wang, D. Guo, X. Qu
  • Computer Science
    IEEE transactions on neural networks and learning systems
  • 2022
A DL architecture for spectra reconstruction from undersampled data, called MoDern, which follows the iterative reconstruction in solving a sparse model to build the neural network and elaborately design a learnable soft-thresholding to adaptively eliminate the spectrum artifacts introduced by undersampling.
Structured Low-Rank Algorithms: Theory, Magnetic Resonance Applications, and Links to Machine Learning
This article provides a detailed review of recent advances in the recovery of continuous-domain multidimensional signals from their few nonuniform (multichannel) measurements using structured low-rank (SLR) matrix completion formulation, and demonstrates the utility of the framework in a wide range of MR imaging applications.


A Fixed Point Iterative Method for Low $n$-Rank Tensor Pursuit
This paper adapt operator splitting technique and convex relaxation technique to transform the original problem into a convex, unconstrained optimization problem and propose a fixed point iterative method to solve it, called FP-LRTC (Fixed Point for Low Tensor Completion).
Robust recovery of complex exponential signals from random Gaussian projections via low rank Hankel matrix reconstruction
Scalable Tensor Factorizations for Incomplete Data
On Tensor Completion via Nuclear Norm Minimization
A convex optimization approach to tensor completion is investigated by directly minimizing a tensor nuclear norm and it is proved that this leads to an improved sample size requirement and a series of algebraic and probabilistic techniques, which may be of independent interests and could be useful in other tensor-related problems.
Trace Norm Regularized CANDECOMP/PARAFAC Decomposition With Missing Data
A novel trace norm regularized CANDECOMP/PARAFAC decomposition (TNCP) method for simultaneous tensor decomposition and completion that is significantly faster than the state-of-the-art methods and scales to larger problems.
Tensor completion for estimating missing values in visual data
An algorithm to estimate missing values in tensors of visual data by laying out the theoretical foundations and building a working algorithm is proposed, which is more accurate and robust than heuristic approaches.
Simultaneous Tensor Decomposition and Completion Using Factor Priors
This paper proposes a method called simultaneous tensor decomposition and completion (STDC) that combines a rank minimization technique with Tucker model decomposition, and uses factor priors, which are usually known a priori in real-world tensor objects, to characterize the underlying joint-manifold drawn from the model factors.
Nuclear Norms for Tensors and Their Use for Convex Multilinear Estimation
A proper extension of the concept of Shatten-q norms for matrices — including the nuclear norm — to higher order tensors is introduced and a general class of nonsmooth convex optimization problems where the tensorial nuclear norm is employed as a penalty function is studied.
The Convex Geometry of Linear Inverse Problems
This paper provides a general framework to convert notions of simplicity into convex penalty functions, resulting in convex optimization solutions to linear, underdetermined inverse problems.
Robust Spectral Compressed Sensing via Structured Matrix Completion
This paper develops a novel algorithm, called enhanced matrix completion (EMaC), based on structured matrix completion that does not require prior knowledge of the model order to recover a spectrally sparse signal from a small random subset of its n time domain samples.