Multi-view low-rank sparse subspace clustering

By Maria Brbic, Ivica Kopriva. Pattern Recognition, 2017

453 citations recorded by the source corpus.

Abstract Most existing approaches address multi-view subspace clustering problem by constructing the affinity matrix on each view separately and afterwards propose how to extend spectral clustering algorithm to handle multi-view data. This paper presents an approach to multi-view subspace clustering that learns a joint subspace representation by constructing affinity matrix shared among all views. Relying on the importance of both low-rank and sparsity constraints in the construction of the affinity matrix, we introduce the objective that balances between the agreement across different views, while at the same time encourages sparsity and low-rankness of the solution. Related low-rank and sparsity constrained optimization problem is for each view solved using the alternating direction method of multipliers. Furthermore, we extend our approach to cluster data drawn from nonlinear subspaces by solving the corresponding problem in a reproducing kernel Hilbert space. The proposed algorithm outperforms state-of-the-art multi-view subspace clustering algorithms on one synthetic and four real-world datasets.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/5865032c5b157975dd5e412c7336ac332f386ad7

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

1. Nino Antulov-Fantulin
2. Ivica Kopriva
3. Hadi Kazemi
4. Dijana Tolic
5. Fariborz Taherkhani
6. Ali Dabouei
7. Jeremy M. Dawson
8. Nasser M. Nasrabadi

In Computer Vision and Pattern Recognition

1. En Zhu
2. Wenxuan Tu
3. Nino Antulov-Fantulin
4. Xinwang Liu
5. Pei Zhang
6. Ivica Kopriva
7. Sihang Zhou
8. Hadi Kazemi

In Media Technology

1. En Zhu
2. Xinwang Liu
3. Pei Zhang
4. Ivica Kopriva
5. Nasser M. Nasrabadi

Citations in the corpus

• Structured Graph Learning for Scalable Subspace Clustering: From Single View to Multiview (2021, unknown): “For instance, multi-view low-rank sparse subspace clustering (MLRSSC) [59] harnessing both low-rank and sparsity constraints shows much better performance than previous methods, such as [60].”
• Multi-view Contrastive Graph Clustering (2021, unknown): “Multi-view low-rank sparse subspace clustering [Brbi´c and Kopriva, 2018] obtains a joint subspace representation across all views by learning an affinity matrix constrained by sparsity and low-rank constraint.”
• A Weakly Supervised Fine Label Classifier Enhanced by Coarse Supervision (2019, background): “The low rank and sparse representation of the high dimensional data is based on the fact that a correlation often exists among the data which belong to the same class such that a low rank subspace captures the global and smooth structures, while a sparse subspace captures the local structures and fi”
• A Nonlinear Orthogonal Non-Negative Matrix Factorization Approach to Subspace Clustering (2017, background): “Joint low-rank sparseness 15 constrained approach can be extended to multi-view clustering [15].”
• Scalable Multi-view Subspace Clustering with Unified Anchors (2021, methodology): “Multi-view Low-rank Sparse Subspace Clustering (MLRSSC) [2] learns subspace representations by constructing affinity matrices shared among views and solves the associated low-rank and sparse constrained problems.”
• Essential Tensor Learning for Multi-View Spectral Clustering (2018, unknown): “For example, [29] combines the advantages of both LRR and SSC.”