English

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

Computer Vision and Pattern Recognition 2017-07-11 v4

Abstract

This paper focus on recovering multi-dimensional data called tensor from randomly corrupted incomplete observation. Inspired by reweighted l1l_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.

Keywords

Cite

@article{arxiv.1611.05964,
  title  = {Reweighted Low-Rank Tensor Completion and its Applications in Video Recovery},
  author = {Baburaj M. and Sudhish N. George},
  journal= {arXiv preprint arXiv:1611.05964},
  year   = {2017}
}

Comments

Algorithm 1 is inefficient since line 2 is processed n 3 times need to be changed There are inconsistent notations throughout the manuscript Unitary Tensor are not defined