基于稀疏与低秩联合表示的多样性多视图子空间聚类

doi:10.3778/j.issn.1002-8331.2406-0365

摘要/Abstract

摘要： 现有的大多数多视图子空间聚类（multi-view subspace clustering，MVSC）算法着重于视图之间的一致性，却忽略了视图之间的多样性，同时也未充分考虑多视图局部结构中的丰富信息，导致多视图信息挖掘不充分，聚类结果不理想。针对这一问题，提出了一种基于稀疏与低秩联合表示的多样性多视图子空间聚类算法。通过共同自表示，视图间的一致性和多样性得以抽取。在实验中，与八种算法进行了比较，并在五个数据集上进行了评估。评价指标包括归一化互信息（normalized mutual information，NMI）、调整兰德指数（adjusted Rand index，ARI）、准确率（accuracy，ACC）和纯度（purity，PUR）。在五个数据集上，提出算法的效果在四个指标上均属于最优。此外，时间成本分析表明该算法与其他算法具有相当的水平，而消融实验证明了视图间一致性和多样性刻画方式的重要性，且该算法在五个数据集上均具备良好的收敛性。实验结果表明，相比于现存相关方法，该方法展现出一定的优越性。

关键词: 多视图聚类, 一致性与多样性, 稀疏表示, 低秩表示, 流形结构

Abstract: Most multi-view subspace clustering methods only consider the consistency among multi-views, thus ignoring the diversity among multi-views. At the same time, it does not take into account the rich information in the local structure of multi-views, the sparse and low-rank characteristics of multi-views, and the cross-view commonality and inconsistency in subspace representations, resulting in insufficient multi-view information mining and unsatisfactory clustering results. For this purpose, this paper proposes a diversity-aware multi-view subspace clustering based on sparse and low-rank joint representation algorithm. Through common self representation, consistency and diversity between views can be extracted. In the experiment, eight algorithms are compared and evaluated on five datasets. The evaluation indicators include normalized mutual information (NMI), adjusted Rand index (ARI), accuracy (ACC), and purity (PUR). On the five datasets, the performance of the algorithm proposed in this paper is optimal on all four indicators. In addition, the analysis of time costs indicates that the algorithm proposed in this study has a comparable level of computational cost compared to other algorithms. Ablation experiments have demonstrated the importance of depicting consistency and diversity between views, and this algorithm exhibits good convergence on all five datasets. The experimental results show that compared to existing relevant methods, this method exhibits certain advantages.

Key words: multi-view clustering, consistency and diversity, sparse constraints, low-rank constraint, manifold structure

赵悦, 胡良臣, 杨影, 阮芷萱, 杜同春, 接标, 罗永龙. 基于稀疏与低秩联合表示的多样性多视图子空间聚类[J]. 计算机工程与应用, 2025, 61(20): 206-217.

ZHAO Yue, HU Liangchen, YANG Ying, RUAN Zhixuan, DU Tongchun, JIE Biao, LUO Yonglong. Diversity-Aware Multi-View Subspace Clustering Based on Sparse and Low-Rank Joint Representation[J]. Computer Engineering and Applications, 2025, 61(20): 206-217.

参考文献

[1] CHAO G, SUN S, BI J. A survey on multi-view clustering[J]. IEEE Transactions on Artificial Intelligence, 2021, 2(2): 146-168.
[2] 王晶. 多视图子空间聚类方法研究[D]. 绵阳: 西南科技大学, 2024.
WANG J. Research on multi-view subspace clustering methods[D]. Mianyang: Southwest University of Science and Technology, 2024.
[3] GAO H C, NIE F P, LI X L, et al. Multi-view subspace clustering[C]//Proceedings of the IEEE International Conference on Computer Vision. Piscataway: IEEE, 2015: 4238-4246.
[4] WANG Y, ZHANG W J, WU L, et al. Iterative views agreement: an iterative low-rank based structured optimization method to multi-view spectral clustering[C]//Proceedings of the 25th International Joint Conference on Artificial Intelligence. New York: AAAI, 2016: 2153-2159.
[5] ZHANG C Q, HU Q H, FU H Z, et al. Latent multi-view subspace clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 4333-4341.
[6] BRBI? M, KOPRIVA I. Multi-view low-rank sparse subspace clustering[J]. Pattern Recognition, 2018, 73: 247-258.
[7] LYU J C, KANG Z, WANG B Y, et al. Multi-view subspace clustering via partition fusion[J]. Information Sciences, 2021, 560: 410-423.
[8] WANG X B, GUO X J, LEI Z, et al. Exclusivity-consistency regularized multi-view subspace clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 1-9.
[9] TANG C, ZHU X Z, LIU X W, et al. Learning a joint affinity graph for multiview subspace clustering[J]. IEEE Transactions on Multimedia, 2019, 21(7): 1724-1736.
[10] ZHANG C, FU H, HU Q, et al. Generalized latent multi-view subspace clustering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(1): 86-99.
[11] CAO X C, ZHANG C Q, FU H Z, et al. Diversity-induced multi-view subspace clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2015: 586-594.
[12] 刘浩翰, 杜嘉欣, 李建伏. 两级联合融合的多视图子空间聚类改进算法[J]. 计算机应用与软件, 2023, 40(12): 299-304.
LIU H H, DU J X, LI J F. Improved multi-view subspace clustering algorithm based on two-level joint fusion[J]. Computer Applications and Software, 2023, 40(12): 299-304.
[13] CAI X S, HUANG D, ZHANG G Y, et al. Seeking commonness and inconsistencies: a jointly smoothed approach to multi-view subspace clustering[J]. Information Fusion, 2023, 91: 364-375.
[14] 吴杰, 万源, 刘秋杰. 一致块对角和限定的多视角子空间聚类算法[J]. 计算机科学, 2025, 52(4): 138-146.
WU J, WAN Y, LIU Q J. Multi-view subspace clustering algorithm limited by diagonal sum of uniform blocks[J]. Computer Science, 2025, 52(4): 138-146.
[15] ABAVISANI M, PATEL V M. Deep multimodal subspace clustering networks[J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(6): 1601-1614.
[16] CUI C, REN Y, PU J, et al. Deep multi-view subspace clustering with anchor graph[J]. arXiv:2305.06939, 2023.
[17] ELHAMIFAR E, VIDAL R. Sparse subspace clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2009: 2790-2797.
[18] VIDAL R, FAVARO P. Low rank subspace clustering (LRSC)[J]. Pattern Recognition Letters, 2014, 43: 47-61.
[19] NIE F P, CAI G H, LI J, et al. Auto-weighted multi-view learning for image clustering and semi-supervised classification[J]. IEEE Transactions on Image Processing, 2018, 27(3): 1501-1511.
[20] WANG S W, LIU X W, ZHU X Z, et al. Fast parameter-free multi-view subspace clustering with consensus anchor guidance[J]. IEEE Transactions on Image Processing, 2022, 31: 556-568.
[21] LU T X. Solution of the matrix equation AX?XB=C[J]. Computing, 1986, 37(4): 351-355.
[22] CAI J F, CANDèS E J, SHEN Z W. A singular value thresholding algorithm for matrix completion[J]. SIAM Journal on Optimization, 2010, 20(4): 1956-1982.
[23] WRIGHT S J, NOWAK R D, FIGUEIREDO M A T. Sparse reconstruction by separable approximation[J]. IEEE Transactions on Signal Processing, 2009, 57(7): 2479-2493.
[24] REN Z, SUN Q, WU B, et al. Learning latent low-rank and sparse embedding for robust image feature extraction[J]. IEEE Transactions on Image Processing, 2020, 29(1): 2094-2107.
[25] STREHL A. Cluster ensembles a knowledge reuse framework for combining multiple partitions[J]. The Journal of Machine Learning Research, 2003, 3: 583-617.
[26] HUANG D, WANG C D, PENG H X, et al. Enhanced ensemble clustering via fast propagation of cluster-wise similarities[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2021, 51(1): 508-520.
[27] LIU X, ZHU X, LI M, et al. Late fusion incomplete multi-view clustering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41(10): 2410-2423.
[28] ZHANG Z, LIU L, SHEN F M, et al. Binary multi-view clustering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41(7): 1774-1782.
[29] NG A Y, JORDAN M I, WEISS Y. On spectral clustering: Analysis and an algorithm[C]//Proceedings of the 14th International Conference on Neural Information Processing Systems, 2001: 849-856.
[30] TZORTZIS G, LIKAS A. Kernel-based weighted multi-view clustering[C]//Proceedings of the IEEE 12th International Conference on Data Mining. Piscataway: IEEE, 2012: 675-684.
[31] XIA R K, PAN Y, DU L, et al. Robust multi-view spectral clustering via low-rank and sparse decomposition[J]. IEEE Transactions on Cybernetics, 2022, 52(4): 2467-2476.
[32] LI Y Q, NIE F P, HUANG H, et al. Large-scale multi-view spectral clustering via bipartite graph[J]. Proceedings of the 29th AAAI Conference on Artificial Intelligence, 2015: 2750-2756.
[33] ZHAN K, ZHANG C Q, GUAN J P, et al. Graph learning for multiview clustering[J]. IEEE Transactions on Cybernetics, 2018, 48(10): 2887-2895.
[34] SUN M J, ZHANG P, WANG S W, et al. Scalable multi-view subspace clustering with unified anchors[C]//Proceedings of the 29th ACM International Conference on Multimedia, 2021: 3528-3536.