Non-Convex Approaches for Low-Rank Tensor Completion under Tubal Sampling
Non-Convex Approaches for Low-Rank Tensor Completion under Tubal Sampling
复制标题
DOI：
10.1109/icassp49357.2023.10094847
复制
发表时间：
2023-03
期刊：
影响因子：
--
通讯作者：
Zheng Tan;Longxiu Huang;HanQin Cai;Y. Lou
中科院分区：
文献类型：
--
作者：
Zheng Tan;Longxiu Huang;HanQin Cai;Y. Lou
张量补全是现代数据分析中的一个重要问题。在这项工作中，我们研究了一种特定的采样策略，称为输卵管采样。我们提出了两种易于实现的新型非凸张量补全框架，名为张量 L1-L2 (TL12) 和通过 CUR 的张量补全 (TCCUR)。我们测试了两种方法在合成数据和彩色图像修复问题上的效率。经验结果揭示了在低采样率下这两种方法的准确性和时间效率之间的权衡。它们中的每一种都至少在一个方面优于一些经典的完成方法。
Tensor completion is an important problem in modern data analysis. In this work, we investigate a specific sampling strategy, referred to as tubal sampling. We propose two novel non-convex tensor completion frameworks that are easy to implement, named tensor L1-L2 (TL12) and tensor completion via CUR (TCCUR). We test the efficiency of both methods on synthetic data and a color image inpainting problem. Empirical results reveal a trade-off between the accuracy and time efficiency of these two methods in a low sampling ratio. Each of them outperforms some classical completion methods in at least one aspect.