Modewise Operators, the Tensor Restricted Isometry Property, and Low-Rank Tensor Recovery

TL;DR

This paper introduces modewise measurement operators for low-rank tensor recovery that are memory-efficient, satisfy the tensor restricted isometry property, and enable accurate reconstruction from fewer measurements.

Contribution

It proposes modewise measurement schemes for tensors that are computationally efficient, satisfy the tensor RIP, and outperform traditional vectorized approaches.

Findings

Modewise operators require less memory than traditional methods.

They provably satisfy the tensor restricted isometry property.

Experimental results show successful tensor recovery with fewer measurements.

Abstract

Recovery of sparse vectors and low-rank matrices from a small number of linear measurements is well-known to be possible under various model assumptions on the measurements. The key requirement on the measurement matrices is typically the restricted isometry property, that is, approximate orthonormality when acting on the subspace to be recovered. Among the most widely used random matrix measurement models are (a) independent sub-gaussian models and (b) randomized Fourier-based models, allowing for the efficient computation of the measurements. For the now ubiquitous tensor data, direct application of the known recovery algorithms to the vectorized or matricized tensor is awkward and memory-heavy because of the huge measurement matrices to be constructed and stored. In this paper, we propose modewise measurement schemes based on sub-gaussian and randomized Fourier measurements. These modewise operators act on the pairs or other small subsets of the tensor modes separately. They require significantly less memory than the measurements working on the vectorized tensor, provably satisfy the tensor restricted isometry property and experimentally can recover the tensor data from fewer measurements and do not require impractical storage.