Abstract:Benefiting from global rank constraints, the low-rank representation (LRR) method has been shown to be an effective solution to subspace learning. However, the global mechanism also means that the LRR model is not suitable for handling large-scale data or dynamic data. For large-scale data, the LRR method suffers from high time complexity, and for dynamic data, it has to recompute a complex rank minimization for the entire data set whenever new samples are dynamically added, making it prohibitively expensive. Existing attempts to online LRR either take a stochastic approach or build the representation purely based on a small sample set and treat new input as out-of-sample data. The former often requires multiple runs for good performance and thus takes longer time to run, and the latter formulates online LRR as an out-of-sample classification problem and is less robust to noise. In this paper, a novel online LRR subspace learning method is proposed for both large-scale and dynamic data. The proposed algorithm is composed of two stages: static learning and dynamic updating. In the first stage, the subspace structure is learned from a small number of data samples. In the second stage, the intrinsic principal components of the entire data set are computed incrementally by utilizing the learned subspace structure, and the LRR matrix can also be incrementally solved by an efficient online singular value decomposition algorithm. The time complexity is reduced dramatically for large-scale data, and repeated computation is avoided for dynamic problems. We further perform theoretical analysis comparing the proposed online algorithm with the batch LRR method. Finally, experimental results on typical tasks of subspace recovery and subspace clustering show that the proposed algorithm performs comparably or better than batch methods, including the batch LRR, and significantly outperforms state-of-the-art online methods.

Fast Universal Low Rank Representation

FaLRR: A Fast Low Rank Representation Solver

Bilateral Fast Low-Rank Representation With Equivalent Transformation for Subspace Clustering

LatLRR for subspace clustering via reweighted Frobenius norm minimization

Online Low-Rank Representation Learning for Joint Multi-Subspace Recovery and Clustering

Laplacian Regularized Low-Rank Representation and Its Applications

Robust Subspace Segmentation by Low-Rank Representation.

Analysis and Improvement of Low Rank Representation for Subspace Segmentation

Improving Generalization of Double Low-Rank Representation Using Schatten-p Norm

Constrained Low-Rank Learning Using Least Squares-Based Regularization

Robust Latent Low Rank Representation for Subspace Clustering

Robust low rank representation via feature and sample scaling

Fixed-rank Representation for Unsupervised Visual Learning

Sparse fixed-rank representation for robust visual analysis.

Latent Low-Rank Representation With Weighted Distance Penalty for Clustering

Subspace Clustering Based on Latent Low-Rank Representation with Transformed Schatten-1 Penalty Function

Hyperspectral Anomaly Detection Based on Low-Rank Representation Using Local Outlier Factor

Subspace Clustering Based on Latent Low Rank Representation with Weighted Nuclear Norm Minimization.

Fast Low Rank Representation Based Spatial Pyramid Matching for Image Classification.

Fast Low-Rank Subspace Segmentation

Robust sparse low-rank embedding for image dimension reduction