Abstract:Tensor decompositions have been successfully applied to multidimensional data recovery. However, classical tensor decompositions are not suitable for emerging spatio-irregular multidimensional data (i.e., spatio-irregular tensor), whose spatial domain is non-rectangular, e.g., spatial transcriptomics data from bioinformatics and semantic units from computer vision. By using preprocessing (e.g., zero-padding or element-wise 0-1 weighting), the spatio-irregular tensor can be converted to a spatio-regular tensor and then classical tensor decompositions can be applied, but this strategy inevitably introduces bias information, leading to artifacts. How to design a tensor-based method suitable for emerging spatio-irregular tensors is an imperative challenge. To address this challenge, we propose a learnable transform-assisted tensor singular value decomposition (LTA-TSVD) for spatio-irregular tensor recovery, which allows us to leverage the intrinsic structure behind the spatio-irregular tensor. Specifically, we design a learnable transform to project the original spatio-irregular tensor into its latent spatio-regular tensor, and then the latent low-rank structure is captured by classical TSVD on the resulting regular tensor. Empowered by LTA-TSVD, we develop spatio-irregular low-rank tensor completion (SIR-LRTC) and spatio-irregular tensor robust principal component analysis (SIR-TRPCA) models for the spatio-irregular tensor imputation and denoising respectively, and we design corresponding solving algorithms with theoretical convergence. Extensive experiments including the spatial transcriptomics data imputation and hyperspectral image denoising show SIR-LRTC and SIR-TRPCA are superior performance to competing approaches and benefit downstream applications.

Low-Tubal-Rank Tensor Recovery with Multilayer Subspace Prior Learning

Guaranteed Tensor Recovery Fused Low-rankness and Smoothness

Tensor Nonconvex Unified Prior For Tensor Recovery

Robust Low-Rank Tensor Recovery Via Nonconvex Singular Value Minimization.

Robust Low-Rank Tensor Recovery with Rectification and Alignment

Learnable Transform-Assisted Tensor Decomposition for Spatio-Irregular Multidimensional Data Recovery

Tensor N-tubal rank and its convex relaxation for low-rank tensor recovery

Unsupervised learning low-rank tensor from incomplete and grossly corrupted data

A Scale-Invariant Relaxation in Low-Rank Tensor Recovery with an Application to Tensor Completion

Trainable Subspaces for Low Rank Tensor Completion: Model and Analysis

Tensor Recovery via Nonconvex Low-Rank Approximation

Low rank tensor completion with sparse regularization in a transformed domain

A robust low‐rank tensor completion model with sparse noise for higher‐order data recovery

Multiplex Transformed Tensor Decomposition for Multidimensional Image Recovery.

Multilayer Collaborative Low-Rank Coding Network for Robust Deep Subspace Discovery

T-SVD Based Non-convex Tensor Completion and Robust Principal Component Analysis.

Robust low tubal rank tensor recovery via L 2 E criterion

A Novel Tensor Factorization-Based Method with Robustness to Inaccurate Rank Estimation

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

A Novel Sparsity Measure for Tensor Recovery

Low-Tubal-Rank Tensor Recovery via Factorized Gradient Descent