Abstract:Since higher-order tensors are naturally suitable for representing multi-dimensional data in real-world, e.g., color images and videos, low-rank tensor representation has become one of the emerging areas in machine learning and computer vision. However, classical low-rank tensor representations can only represent data on finite meshgrid due to their intrinsical discrete nature, which hinders their potential applicability in many scenarios beyond meshgrid. To break this barrier, we propose a low-rank tensor function representation (LRTFR), which can continuously represent data beyond meshgrid with infinite resolution. Specifically, the suggested tensor function, which maps an arbitrary coordinate to the corresponding value, can continuously represent data in an infinite real space. Parallel to discrete tensors, we develop two fundamental concepts for tensor functions, i.e., the tensor function rank and low-rank tensor function factorization. We theoretically justify that both low-rank and smooth regularizations are harmoniously unified in the LRTFR, which leads to high effectiveness and efficiency for data continuous representation. Extensive multi-dimensional data recovery applications arising from image processing (image inpainting and denoising), machine learning (hyperparameter optimization), and computer graphics (point cloud upsampling) substantiate the superiority and versatility of our method as compared with state-of-the-art methods. Especially, the experiments beyond the original meshgrid resolution (hyperparameter optimization) or even beyond meshgrid (point cloud upsampling) validate the favorable performances of our method for continuous representation.

Multi-dimensional Signal Recovery using Low-rank Deconvolution

Fast Unsupervised Tensor Restoration via Low-rank Deconvolution

Nonconvex Regularization Based Sparse Recovery and Demixing with Application to Color Image Inpainting

A reciprocity-aware, low-rank regularization for multidimensional deconvolution

Low-Rank Plus Sparse Reconstruction Using Dictionary Learning for 3D-Mri

Sparsity-Enhanced Multilayered Non-Convex Regularization with Epigraphical Relaxation for Debiased Signal Recovery

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

Adaptive Structure-constrained Robust Latent Low-Rank Coding for Image Recovery

A Nonconvex Approach to Low-Rank and Sparse Matrix Decomposition with Application to Video Surveillance.

Tensor Recovery via Nonconvex Low-Rank Approximation

Structured Low-Rank Algorithms: Theory, Magnetic Resonance Applications, and Links to Machine Learning

Recovery of damped exponentials using structured low rank matrix completion

Robust and Fast Low-Rank Deep Convolutional Feature Recovery: Toward Information Retention and Accelerated Convergence

Fully Decomposable Compressive Sampling with Joint Optimization for Multidimensional Sparse Representation

A physics-aware, low-rank regularization for multidimensional deconvolution

Robust auto-weighted projective low-rank and sparse recovery for visual representation

Multi-Dimensional Visual Data Completion via Low-Rank Tensor Representation Under Coupled Transform

Unrolled Compressed Blind-Deconvolution

Multi-Sparse Signal Recovery for Compressive Sensing

Low-Rank Tensor Function Representation for Multi-Dimensional Data Recovery