Journal
IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING
Volume 12, Issue 6, Pages 1405-1419Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSTSP.2018.2879185
Keywords
Tensor Schatten-p norm; low-rank; tensor decomposition; convex optimization
Categories
Funding
- National Basic Research Program of China (973 Program) [2015CB352502]
- National Natural Science Foundation of China [61625301, 61731018]
- Microsoft Research Asia
Ask authors/readers for more resources
In this paper, we propose a new definition of tensor Schatten-p norm (t-Schatten-p norm) based on t-SVD, and prove that this norm has similar properties to matrix Schatten-p norm. More importantly, the t-Schatten-p norm can better approximate the l(1) norm of the tensor multi-rank with 0 < p < 1. Therefore, it can be used for the Low-Rank Tensor Recovery problems as a tighter regularizer. We further prove the tensor multi-Schatten-p norm surrogate theorem and give an efficient algorithm accordingly. By decomposing the target tensor into many small-scale tensors, the non-convex optimization problem (0 < p < 1) is transformed into many convex sub-problems equivalently, which can greatly improve the computational efficiency when dealing with large-scale tensors. Finally, we provide the theories on the conditions for exact recovery in the noiseless case and give the corresponding error bounds for the noise case. Experimental results on both synthetic and real-world datasets demonstrate the superiority of our t-Schattern-p norm in the Tensor Robust Principle Component Analysisand the Tensor Completion problems.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available