Abstract | ||
---|---|---|
•The robust block tensor principle component analysis is proposed to extract the low-rank and sparse components in block tensors for a good analysis scale.•The alternating direction method of multipliers can divide the problem into two main problems which can be solved efficiently by a proposed iterative block tensor singular value soft thresholding and classical iterative soft thresholding.•Experiments on image denoising and shadow removal demonstrate the enhanced performance in comparison with classical PCA methods. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1016/j.sigpro.2019.107271 | Signal Processing |
Keywords | Field | DocType |
Tensor singular value decomposition,Tensor nuclear norm,Block tensor,Robust tensor principal component analysis | Noise reduction,Singular value decomposition,Mathematical optimization,Normalization (statistics),Singular value,Tensor,Algorithm,Matrix norm,Mathematics,Principal component analysis,Color image | Journal |
Volume | ISSN | Citations |
166 | 0165-1684 | 4 |
PageRank | References | Authors |
0.38 | 0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lanlan Feng | 1 | 4 | 0.38 |
Yipeng Liu | 2 | 43 | 5.93 |
Longxi Chen | 3 | 27 | 2.34 |
Xiang Zhang | 4 | 14 | 1.54 |
Ce Zhu | 5 | 1473 | 117.79 |