DLRF-Net: A Progressive Deep Latent Low-Rank Fusion Network for Hierarchical Subspace Discovery
Authors: 
Zhao Zhang, Jiahuan Ren, Haijun Zhang, Zheng Zhang, Guangcan Liu, 
Shuicheng YanAuthors Info & Claims
Zhao Zhang, Jiahuan Ren, Haijun Zhang, Zheng Zhang, Guangcan Liu, Shuicheng YanAuthors Info & ClaimsArticle No.: 5, Pages 1 - 24
Abstract
Low-rank coding-based representation learning is powerful for discovering and recovering the subspace structures in data, which has obtained an impressive performance; however, it still cannot obtain deep hidden information due to the essence of single-layer structures. In this article, we investigate the deep low-rank representation of images in a progressive way by presenting a novel strategy that can extend existing single-layer latent low-rank models into multiple layers. Technically, we propose a new progressive Deep Latent Low-Rank Fusion Network (DLRF-Net) to uncover deep features and the clustering structures embedded in latent subspaces. The basic idea of DLRF-Net is to progressively refine the principal and salient features in each layer from previous layers by fusing the clustering and projective subspaces, respectively, which can potentially learn more accurate features and subspaces. To obtain deep hidden information, DLRF-Net inputs shallow features from the last layer into subsequent layers. Then, it aims at recovering the hierarchical information and deeper features by respectively congregating the subspaces in each layer of the network. As such, one can also ensure the representation learning of deeper layers to remove the noise and discover the underlying clean subspaces, which will be verified by simulations. It is noteworthy that the framework of our DLRF-Net is general and is applicable to most existing latent low-rank representation models, i.e., existing latent low-rank models can be easily extended to the multilayer scenario using DLRF-Net. Extensive results on real databases show that our framework can deliver enhanced performance over other related techniques.
References
[1]
Y. Bengio, A. Courville, and P. Vincent. 2013. Representation learning: A review and new perspectives. IEEE Trans. Pattern Anal. Mach. Intell 35, 8 (2013), 1798–1828.
[2]
J. Cunningham and Z. Ghahramani. 2015. Linear dimensionality reduction: Survey, insights, and generalizations. J. Mach. Learn. Res. 16, 1 (2015), 2859–2900.
[3]
F. Wang and J. Sun. 2015. Survey on distance metric learning and dimensionality reduction in data mining. Data Min. Knowl. Discov 29, 2 (2015), 534–564.
[4]
J. Wright, Y. Ma, and J. Mairal. 2010. Sparse representation for computer vision and pattern recognition. Proc. IEEE 98, 6 (2010), 1031–1044.
[5]
G. Liu, Z. Lin, and S. C. Yan. 2013. Robust recovery of subspace structures by low-rank representation. IEEE Trans. Pattern Anal. Mach. Intell 35, 1 (2013), 171–184. 2012.88
[6]
F. Fomin, P. Golovach, and F. Panolan. 2020. Parameterized low-rank binary matrix approximation. Data Min. Knowl. Discov 34, 2 (2020), 478–532.
[7]
E. J. Candès, X. Li, and Y. Ma. 2011. Robust principal component analysis?. J. ACM 58, 3 (2011), 1–11.
[8]
B. K. Bao, G. C. Liu, and C. Xu. 2012. Inductive robust principal component analysis. IEEE Trans. Image Process. 21, 8 (2012), 3794–3800.
[9]
G. C. Liu and S. C. Yan. 2011. Latent low-rank representation for subspace segmentation and feature extraction. In Proceedings of the IEEE Conference on Computer Vision. 1615–1622.
[10]
Z. Zhang, S. C. Yan, and M. B. Zhao. 2014. Similarity preserving low-rank representation for enhanced data representation and effective subspace learning. Neural Networks 53 (2014), 81–94. org/10.1016/j.neunet.2014.01.001
[11]
Z. Zhang, F. Li, M. Zhao, L. Zhang, and S. Yan. 2016. Joint low-rank and sparse principal feature coding for enhanced robust representation and visual classification. IEEE Trans. Image Process. 25, 6 (2016), 2429–2443.
[12]
S. Yu and Y. Wu. 2018. Subspace clustering based on latent low rank representation with Frobenius norm minimization. Neurocomputing 275 (2018), 2479–2489.
[13]
Z. Lin, M. Chen, L. Wu, and Y. Ma. 2009. The augmented Lagrange multiplier method for exact recovery of corrupted low-rank matrices. Technical Report UILU-ENG-09-2215. 2012.10.010
[14]
H. Zhang, Z. C. Lin, C. Zhang, and J. Gao. 2014. Robust latent low rank representation for subspace clustering. Neurocomputing 145 (2014), 369–373.
[15]
J. Ren, Z. Zhang, S. Li, Y. Wang, G. Liu, S. Yan, and M. Wang. 2019. Learning hybrid representation by robust dictionary learning in factorized compressed space. IEEE Trans. Image Process. 29, 1 (2020), 3941–3956.
[16]
P. He, X. Xu, J. Ding, and B. Fan. 2020. Low-rank nonnegative matrix factorization on Stiefel manifold. Inf. Sci. 514 (2020), 131–148.
[17]
C. Hou, F. P. Nie, X. L. Li, D.Y. Yi, and Yi Wu. 2014. Joint embedding learning and sparse regression: A framework for unsupervised feature selection. IEEE Trans. Cybern. 44, 6 (2014), 793–804.
[18]
L. Ma, C. H. Wang, B. H. Xiao, and W. Zhou. 2012. Sparse representation for face recognition based on discriminative low-rank dictionary learning. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2586–2593.
[19]
L. Y. Li, S. Li, and Y. Fu. 2014. Learning low-rank and discriminative dictionary for image classification. Image Vis. Comput. 32, 10 (2014), 814–823.
[20]
Y. Zhang, Z. L. Jiang, and L. Davis. 2013. Learning structured low-rank representations for image classification. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition 2013, 676–683.
[21]
J. Chen and Z. Yi. 2014. Sparse representation for face recognition by discriminative low-rank matrix recovery. J. Vis. Commun. Image R 25, 5 (2014), 763–773.
[22]
J. Xie, J. Yang, J. Qian, Y. Tai, and H. Zhang. 2017. Robust nuclear norm-based matrix regression with applications to robust face recognition. IEEE Trans. Image Process. 26, 5 (2017), 2286–2295.
[23]
Z. Xue, J. Du, D. Du, and S. Lyu. 2019. Deep low-rank subspace ensemble for multi-view clustering. Inf. Sci. 482 (2019), 210–227.
[24]
Z. Li and J. Tang. 2017. Weakly-supervised deep nonnegative low-rank model for social image tag refinement and assignment. In Proceedings of AAAI Conference on Artificial Intelligence. 4154–4160.
[25]
J. Zhao, Y. Lv, Z. Zhou, and F. Cao. 2017. A novel deep learning algorithm for incomplete face recognition: Low-rank-recovery network. Neural Networks 94 (2017), 115–124.
[26]
J. Cai, E. Candès, and Z. Shen. 2010. A singular value thresholding algorithm for matrix completion. SIAM J. Opt. 20, 4 (2010), 1956–1982.
[27]
J. Shi and J. Malik. 2000. Normalized cuts and image segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 22, 8 (2000), 888–905.
[28]
S. A. Nene, S. K. Nayar, and H. Murase. 1996. Columbia object image library (COIL-20). Technical Report CUCS-005-96.
[29]
L. Wang, B. Wang, Z. Zhang, Q. Yue, L. Fu, G. Liu, and M. Wang. 2019. Robust auto-weighted projective low-rank and sparse recovery for visual representation. Neural Networks 117 (2019), 201–215.
[30]
Z. Zhang, M. Zhao, F. Li, L. Zhang, and S. Yan. 2017. Robust alternating low-rank representation by joint Lp- and L2,p-norm minimization. Neural Networks 96 (2017), 55–70.
[31]
Z. Zhang, J. Ren, S. Li, R. Hong, Z. Zha, and M. Wang. 2019. Robust subspace discovery by block-diagonal adaptive locality-constrained representation. In Proceedings of the ACM International Conference on Multimedia. 1569–1577.
[32]
J. Hull. 1994. A database for handwritten text recognition research. IEEE Trans. Pattern Anal. Mach. Intell. 16, 5 (1994), 550–554.
[33]
Z. Song, K. Cui, and G. Cheng. 2020. Image set face recognition based on extended low rank recovery and collaborative representation. Int. J. Mach. Learn. Cybern. 11, 1 (2020), 71–80.
[34]
Z. Zhang, L.Wang, S. Li, Y. Wang, Z. Zhang, Z. Zha, and M. Wang. 2019. Adaptive structure-constrained robust latent low-rank coding for image recovery. In Proceedings of the 19th IEEE International Conference on Data Mining.
[35]
Y. Lu, Z. Lai, W. Wong, and X. Li. 2020. Low-rank discriminative regression learning for image classification. Neural Networks 125 (2020), 245–257.
[36]
H. Li and Z. Lin. 2020. Provable accelerated gradient method for nonconvex low rank optimization. Mach. Learn. 109, 1 (2020), 103–134.
[37]
X. Lan, S. Zhang, P. C. Yuen, and R. Chellappa. 2018. Learning common and feature-specific patterns: A novel multiple-sparse-representation based tracker. IEEE Trans. Image Process. 27, 4 (2018), 2022–2037.
[38]
F. V. Fomin, P. A. Golovach, D. Lokshtanov, F. Panolan, and S. Saurabh. 2020. Approximation schemes for low-rank binary matrix approximation problems. ACM Trans. Algorithms 16, 1 (2020), 12:1–12:39.
[39]
D. Graham and N. Allinson. 1998. Characterizing virtual eigensignatures for general purpose face recognition. In Face Recognition: From Theory to Application (NATO ASI Series F, Computer and Systems Sciences). Springer, Berlin, 446–456.
[40]
A. Georghiades, P. Belhumeur, and D. Kriegman. 2001. From few to many: Illumination cone models for face recognition under variable lighting and pose. IEEE Trans. Pattern Anal. Mach. Intell. 23, 6 (2001), 643–660.
[41]
T. Sim, S. Baker, and M. Bsat. 2003. The CMU pose, illuminlation, and expression database. IEEE Trans. Pattern Anal. Mach. Intell. 25, 12 (2003), 1615–1618.
[42]
B. Weyrauch, J. Huang, B. Heisele, and V. Blanz. 2004. Component-based face recognition with 3D morphable models. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. 85–85.
[43]
J. Li, H. Chang, J. Yang, W. Luo, and Y. Fu. 2017. Visual representation and classification by learning group sparse deep stacking network. IEEE Transactions . Image Process. 27, 1 (2017), 464–476.
[44]
H. Xiao, K. Rasul, and R. Vollgraf. 2017. Fashion-mnist: A novel image dataset for benchmarking machine learning algorithms. arXiv:1708.07747v2.
[45]
J. Li and H. Liu. 2017. Projective low-rank subspace clustering via learning deep encoder. In Proceedings of the International Joint Conferences on Artificial Intelligence. 2145–2151.
[46]
X. Chen, J. Li, Y. Song, F. Li, J. Chen, and K. Yang. 2019. Low-rank tensor completion for image and video recovery via capped nuclear norm. IEEE Access. 7 (2019), 112142–112153.
[47]
F. Su, H. Y. Shang, and J. Y. Wang. 2019. Low-rank deep convolutional neural network for multitask learning. Comput. Intell. Neurosci. 2019, 7410701:1–7410701:10.
[48]
Z. Ding and Y. Fu. 2018. Deep transfer low-rank coding for cross-domain learning. IEEE Trans. Neural Netw. Learning Syst. 30, 6 (2018), 1768–1779.
[49]
C. Lu, X. Peng, and Y. Wei. 2019. Low-rank tensor completion with a new tensor nuclear norm induced by invertible linear transforms. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 5996–6004.
[50]
Y. Chen, X. Xiao, and Y. Zhou. 2020. Low-rank quaternion approximation for color image processing. IEEE Trans. Image Process. 29 (2020), 1426–1439. 319
[51]
A. Acharya, R. Goel, A. Metallinou, and I. S. Dhillon. 2019. Online embedding compression for text classification using low rank matrix factorization. In Proceedings of th eAAAI Conference on Artificial Intelligence. 6196–6203.
[52]
J. Kim, J. Sim, and C. Kim. 2015. Video deraining and desnowing using temporal correlation and low-rank matrix completion. IEEE Trans. Image Process. 24, 9 (2015), 2658–2670. 2015.2428933
[53]
H. Zhang and V. Patel. 2017. Convolutional sparse and low-rank coding-based rain streak removal. In Proceedings of the IEEE Winter Conference on Applications of Computer Vision, 1259–1267.
[54]
B. Kim, H. Kim, K. Kim, S. Kim, and J. Kim. 2019. Learning not to learn: Training deep neural networks with biased data. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 9012–9020.
[55]
Y. Wang, X. Lin, L. Wu, W. Zhang, Q. Zhang, and X. Huang. 2015. Robust subspace clustering for multi-view data by exploiting correlation consensus. IEEE Trans. Image Process. 24, 11 (2015), 3939–3949.
[56]
Y. Wang, L. Wu, X. Lin, and J. Gao. 2018. Multiview spectral clustering via structured low-rank matrix factorization. IEEE Trans. Neural Netw. Learning Syst. 29, 10 (2018), 4833–4843.
[57]
P. Giampouras, A. Rontogiannis, and K. Koutroumbas. 2019. Alternating iteratively reweighted least squares minimization for low-rank matrix factorization. IEEE Trans. Signal Process. 67, 2 (2019), 490–503.
[58]
Z. Zhang, J. Ren, Z. Zhang, and G. Liu. 2020. Deep latent low-rank fusion network for progressive subspace discovery. In Proceedings of the International Joint Conferences on Artificial Intelligence. 1–8.
[59]
Z. Kang, X. Zhao, C. Peng, H. Zhu, J. T. Zhou, X. Peng, W. Chen, and Z. Xu. 2020. Partition level multiview subspace clustering. Neural Networks 122 (2019), 279–288. /j.ne unet.2019.10.010
[60]
Z. Kang, G. Shi, S. Huang, W. Chen, X. Pu, J. T. Zhou, and Z. Xu. 2020. Multi-graph fusion for multi-view spectral clustering. Knowledge-Based Syst. 189 (2020), 105102.
[61]
Z. Kang, H. Pan, S. Hoi, and Z. Xu. 2019. Robust graph learning from noisy data. IEEE Trans. Cybern. 2019, 1–11.
Index Terms
- DLRF-Net: A Progressive Deep Latent Low-Rank Fusion Network for Hierarchical Subspace Discovery
Recommendations
Deep latent low-rank fusion network for progressive subspace discovery
IJCAI'20: Proceedings of the Twenty-Ninth International Joint Conference on Artificial IntelligenceLow-rank representation is powerful for recovering and clustering the subspace structures, but it cannot obtain deep hierarchical information due to the single-layer mode. In this paper, we present a new and effective strategy to extend the single-layer ...
Laplacian regularized deep low-rank subspace clustering network
AbstractSelf-expression-based deep subspace clustering, integrating traditional subspace clustering methods into deep learning paradigm to enhance the representative capacity, has become an important branch in unsupervised learning methods. However, most ...
Learnable low-rank latent dictionary for subspace clustering
Highlights- We propose a unified framework to integrate linear or nonlinear feature extraction and subspace clustering.
Graphical abstractDisplay Omitted
AbstractRecently, Self-Expressive-based Subspace Clustering (SESC) has been widely applied in pattern clustering and machine learning as it aims to learn a representation that can faithfully reflect the correlation between data points. However,...
Comments
Information & Contributors
Information
Published In
January 2021
353 pages
Copyright © 2021 Association for Computing Machinery.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 31 March 2021
Accepted: 01 May 2020
Revised: 01 April 2020
Received: 01 February 2020
Published in TOMM Volume 17, Issue 1s
Check for updates
Author Tags
Qualifiers
- Research-article
- Refereed
Funding Sources
- National Key R&D Program of China
- National Natural Science Foundation of China
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 222Total Downloads
- Downloads (Last 12 months)25
- Downloads (Last 6 weeks)1
Reflects downloads up to 26 Sep 2024
Other Metrics
Citations
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign inFull Access
View options
View or Download as a PDF file.
PDFeReader
View online with eReader.
eReaderHTML Format
View this article in HTML Format.
HTML Format