Fully-connected tensor network decomposition for robust tensor completion problem

DOI10.3934/IPI.2023030MaRDI QIDQ6087983zbMATH OpenOpenAlexFDO

Authors Yun-Yang Liu, Xi-Le Zhao, G. J. Song, Yu-Bang Zheng, Michael Ng, Ting-Zhu Huang

Publication date 13 December 2023

Published in Inverse Problems and Imaging (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2110.08754

zbMATH Keywords

fully-connected tensor network decompositionrobust tensor completionexact recovery guarantee

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Nonconvex programming, global optimization (90C26) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10)

Abstract: The robust tensor completion (RTC) problem, which aims to reconstruct a low-rank tensor from partially observed tensor contaminated by a sparse tensor, has received increasing attention. In this paper, by leveraging the superior expression of the fully-connected tensor network (FCTN) decomposition, we propose a $extbfFCTN$-based $extbfr$obust $extbfc$onvex optimization model (RC-FCTN) for the RTC problem. Then, we rigorously establish the exact recovery guarantee for the RC-FCTN. For solving the constrained optimization model RC-FCTN, we develop an alternating direction method of multipliers (ADMM)-based algorithm, which enjoys the global convergence guarantee. Moreover, we suggest a $extbfFCTN$-based $extbfr$obust $extbfn$on$extbfc$onvex optimization model (RNC-FCTN) for the RTC problem. A proximal alternating minimization (PAM)-based algorithm is developed to solve the proposed RNC-FCTN. Meanwhile, we theoretically derive the convergence of the PAM-based algorithm. Comprehensive numerical experiments in several applications, such as video completion and video background subtraction, demonstrate that proposed methods are superior to several state-of-the-art methods.

Recommendations

• Tensor completion via fully-connected tensor network decomposition with regularized factors
• Robust tensor completion: equivalent surrogates, error bounds, and algorithms
• Nonconvex optimization for robust tensor completion from grossly sparse observations
• Robust low-rank tensor completion via new regularized model with approximate SVD
• Nonlocal robust tensor recovery with nonconvex regularization

Cites work

• scientific article; zbMATH DE number 6474942 (Why is no real title available?)
• A block coordinate descent method for regularized multiconvex optimization with applications to nonnegative tensor factorization and completion
• A new initialization method based on normed statistical spaces in deep networks
• A nonlocal low rank model for Poisson noise removal
• Auto-weighted robust low-rank tensor completion via tensor-train
• Bilinear constraint based ADMM for mixed Poisson-Gaussian noise removal
• Clarke Subgradients of Stratifiable Functions
• Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward-backward splitting, and regularized Gauss-Seidel methods
• Distributed optimization and statistical learning via the alternating direction method of multipliers
• Edge detection with mixed noise based on maximum a posteriori approach
• Efficient Tensor Completion for Color Image and Video Recovery: Low-Rank Tensor Train
• Factorization strategies for third-order tensors
• Guarantees of Riemannian optimization for low rank matrix completion
• Hankel matrix rank minimization with applications to system identification and realization
• Low-rank tensor train for tensor robust principal component analysis
• Most tensor problems are NP-hard
• Multi-dimensional image recovery via fully-connected tensor network decomposition under the learnable transforms
• Nonconvex optimization for robust tensor completion from grossly sparse observations
• Nonlinear transform induced tensor nuclear norm for tensor completion
• Parallel matrix factorization for low-rank tensor completion
• Proximal alternating linearized minimization for nonconvex and nonsmooth problems
• Robust principal component analysis?
• Robust tensor completion using transformed tensor singular value decomposition
• Robust tensor train component analysis
• Tensor train rank minimization with nonlocal self-similarity for tensor completion
• Tensor-Based Formulation and Nuclear Norm Regularization for Multienergy Computed Tomography
• Tensor-train decomposition
• The Power of Convex Relaxation: Near-Optimal Matrix Completion
• Third-order tensors as operators on matrices: a theoretical and computational framework with applications in imaging
• Total variation based tensor decomposition for multi-dimensional data with time dimension.

Cited in

(3)

• A random sampling algorithm for fully-connected tensor network decomposition with applications
• Randomized tensor wheel decomposition
• SVD-based algorithms for fully-connected tensor network decomposition

This page was built for publication: Fully-connected tensor network decomposition for robust tensor completion problem

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6087983)