Reweighted Low-Rank Tensor Completion and its Applications in Video Recovery

TL;DR

This paper introduces a reweighted singular value enhancement method for tensor completion, significantly improving recovery of multi-dimensional data like videos from incomplete and corrupted observations.

Contribution

It proposes a novel reweighted singular value scheme combined with t-SVD for tensor completion, enhancing low-rank recovery performance.

Findings

Outperforms existing tensor completion methods in experiments

Effective in recovering corrupted and incomplete video data

Significantly improves low-rank tensor recovery accuracy

Abstract

This paper focus on recovering multi-dimensional data called tensor from randomly corrupted incomplete observation. Inspired by reweighted $l_1$ norm minimization for sparsity enhancement, this paper proposes a reweighted singular value enhancement scheme to improve tensor low tubular rank in the tensor completion process. An efficient iterative decomposition scheme based on t-SVD is proposed which improves low-rank signal recovery significantly. The effectiveness of the proposed method is established by applying to video completion problem, and experimental results reveal that the algorithm outperforms its counterparts.