Abstract:Tensor train (TT) decomposition represents an $N$-order tensor using $O(N)$ matrices (i.e., factors) of small dimensions, achieved through products among these factors. Due to its compact representation, TT decomposition has found wide applications, including various tensor recovery problems in signal processing and quantum information. In this paper, we study the problem of reconstructing a TT format tensor from measurements that are contaminated by outliers with arbitrary values. Given the vulnerability of smooth formulations to corruptions, we use an $\ell_1$ loss function to enhance robustness against outliers. We first establish the $\ell_1/\ell_2$-restricted isometry property (RIP) for Gaussian measurement operators, demonstrating that the information in the TT format tensor can be preserved using a number of measurements that grows linearly with $N$. We also prove the sharpness property for the $\ell_1$ loss function optimized over TT format tensors. Building on the $\ell_1/\ell_2$-RIP and sharpness property, we then propose two complementary methods to recover the TT format tensor from the corrupted measurements: the projected subgradient method (PSubGM), which optimizes over the entire tensor, and the factorized Riemannian subgradient method (FRSubGM), which optimizes directly over the factors. Compared to PSubGM, the factorized approach FRSubGM significantly reduces the memory cost at the expense of a slightly slower convergence rate. Nevertheless, we show that both methods, with diminishing step sizes, converge linearly to the ground-truth tensor given an appropriate initialization, which can be obtained by a truncated spectral method.

Robust Tensor Recovery Via a Nonconvex Approach with Ket Augmentation and Auto-Weighted Strategy

Tensor Nonconvex Unified Prior For Tensor Recovery

Robust Low-Rank Tensor Recovery with Rectification and Alignment

Handling The Non-Smooth Challenge in Tensor SVD: A Multi-Objective Tensor Recovery Framework

Robust low tubal rank tensor recovery via L 2 E criterion

Tensor Recovery with Weighted Tensor Average Rank.

Tensor Recovery via Multi-linear Augmented Lagrange Multiplier Method

Robust tensor recovery with nonconvex and nonsmooth regularization

Mixture Augmented Lagrange Multiplier Method for Tensor Recovery

Auto-weighted robust low-rank tensor completion via tensor-train

Regularized Kaczmarz Algorithms for Tensor Recovery

Multiscale Feature Tensor Train Rank Minimization for Multidimensional Image Recovery

Tensor recovery using the tensor nuclear norm based on nonconvex and nonlinear transformations

Tensor completion via joint reweighted tensor Q -nuclear norm for visual data recovery

Robust Low-rank Tensor Train Recovery

On The Synergy Between Nonconvex Extensions of The Tensor Nuclear Norm for Tensor Recovery

Robust Tensor CUR Decompositions: Rapid Low-Tucker-Rank Tensor Recovery with Sparse Corruptions

Tensor Recovery from Noisy and Multi-Level Quantized Measurements

TIHN: Tensor Improved Huber Norm for low-rank tensor recovery

Simultaneous Rectification and Alignment via Robust Recovery of Low-rank Tensors.

Low-n-rank tensor recovery based on multi-linear augmented Lagrange multiplier method