Low-rank representation with adaptive dictionary learning for subspace clustering
作者：
Highlights：
• The proposed method learns the low-rank features to construct an affinity matrix.
• A dictionary learning strategy uses an orthonormality constraint in the LRR model.
• The dictionary, adaptively learned from original data, makes ALRR robust to noise.
• The convergence of ALRR is theoretically guaranteed under certain conditions.
• ALRR requires at most three iteration computations for optimization.
摘要
•The proposed method learns the low-rank features to construct an affinity matrix.•A dictionary learning strategy uses an orthonormality constraint in the LRR model.•The dictionary, adaptively learned from original data, makes ALRR robust to noise.•The convergence of ALRR is theoretically guaranteed under certain conditions.•ALRR requires at most three iteration computations for optimization.
论文关键词：Low-rank representation,Dictionary learning,Subspace clustering,Spectral clustering
论文评审过程：Received 4 February 2021, Revised 5 April 2021, Accepted 14 April 2021, Available online 18 April 2021, Version of Record 23 April 2021.
论文官网地址：https://doi.org/10.1016/j.knosys.2021.107053
