Abstract
In image classification, the goal was to decide whether an image belongs to a certain category or not. Multiple features are usually employed to comprehend the contents of images substantially for the improvement of classification accuracy. However, it also brings in some new problems that how to effectively combine multiple features together and how to handle the high-dimensional features from multiple views given the small training set. In this paper, we integrate the large-margin idea into the Gaussian process to discover the latent subspace shared by multiple features. Therefore, our approach inherits all the advantages of Gaussian process and large-margin principle. A probabilistic explanation is provided by Gaussian process to embed multiple features into the shared low-dimensional subspace, which derives a strong discriminative ability from the large-margin principle, and thus, the subsequent classification task can be effectively accomplished. Finally, we demonstrate the advantages of the proposed algorithm on real-world image datasets for discovering discriminative latent subspace and improving the classification performance.
Similar content being viewed by others
References
Akaho, S.: A kernel method for canonical correlation analysis. In: IMPS’2001, (2001)
Bach, F.R., Lanckriet, G.R.G., Jordan, M.I.: Multiple kernel learning, conic duality, and the smo algorithm. In: ICML’2004, (2004)
Bishop, C.M., et al.: Pattern Recognition and Machine Learning, vol. 1. Springer, New York (2006)
Blum, A., Mitchell, T.: Combining labeled and unlabeled data with co-training. In: COLT’1998, (1998)
Chapelle, O., Weston, J., Schölkopf, B.: Cluster kernels for semi-supervised, learning. In: NIPS’2002, (2002)
Chaudhuri, K., Kakade, S.M., Livescu, K., Sridharan, K.: Multi-view clustering via canonical correlation analysis. In: ICML’2009, (2009)
Chen, N., Zhu, J., Xing, E.P.: Predictive subspace learning for multi-view data: a large margin approach. In: NIPS’2010, (2010)
Chen, Ning, Zhu, Jun, Sun, Fuchun: Large-margin predictive latent subspace learning for multiview data analysis. IEEE. Trans. Pattern Anal. Mach. Intell. 34(12), 2365–2378 (2012)
Christoudias, C.M., Urtasun, R., Darrell, T.: Multi-view learning in the presence of view disagreement
Diethe, T., Hardoon, D.R., Shawe-Taylor, J.: Multiview fisher discriminant analysis. In: NIPS workshop on learning from multiple sources (2008)
Guillaumin, M., Verbeek, J., Schmid, C.: Multimodal semi-supervised learning for image classification. In: CVPR’2010, (2010)
Jia, Y., Salzmann, M., Darrell, T.: Factorized latent spaces with structured sparsity. In: NIPS’2010, (2010)
Kakade, S.M., Foster, D.P.: Multi-view regression via canonical correlation analysis. In: Learning Theory, pp. 82–96. Springer (2007)
Lanckriet, G.R.G., Cristianini, N., El Ghaoui, L., Bartlett, P., Jordan, M.I.: Learning the kernel matrix with semi-definite programming. Computer (2002)
Lawrence, D.N.: Gaussian process models for visualisation of high dimensional data (2004)
Lawrence, N.D., Quiñonero-Candela, J.: Local distance preservation in the gp-lvm through back constraints. In: ICML’2006, (2006)
Luo, Y., Tao, D., Xu, C., Liu, H., Wen, Y.: Multiview vector-valued manifold regularization for multilabel image classification. IEEE Trans. Neural Netw. Learn. Syst. 24(709—-722), 2676–2687 (2013)
Memisevic, R.: Kernel information embeddings. In: ICML’2006, (2006)
Memisevic, Roland, Sigal, Leonid, Fleet, David J.: Shared kernel information embedding for discriminative inference. IEEE. Trans. Pattern Anal. Mach. Intell. 34(4), 778–790 (2012)
Møller, Martin Fodslette: A scaled conjugate gradient algorithm for fast supervised learning. Neural Netw. 6(4), 525–533 (1993)
Nigam, K., Ghani, R.: Analyzing the effectiveness and applicability of co-training. In: CIKM’2000, (2000)
Rakotomamonjy, A., Bach, F., Canu, S., Grandvalet, Y.: Simplemkl. J. Mach. Learn. Res. 9, 2491–2521 (2008)
Rasmussen, C.E. Gaussian processes for machine learning (2006)
Shon, A.P., Grochow, K., Hertzmann, A., Rao, R.P.: Learning shared latent structure for image synthesis and robotic imitation. Adv. Neural Inf. Process. Syst. 18, 1233 (2006)
Sigal, L., Memisevic, R., Fleet, D.J.: Shared kernel information embedding for discriminative inference. In: CVPR’2009, (2009)
Sindhwani, V., Niyogi, P., Belkin, M.: A co-regularization approach to semi-supervised learning with multiple views. In: ICML Workshop on Learning with Multiple Views (2005)
Sonnenburg, Sören, Rätsch, Gunnar, Schäfer, Christin, Schölkopf, Bernhard: Large scale multiple kernel learning. J. Mach. Learn. Res. 7, 1531–1565 (2006)
Tao, D., Liu, W.: Multiview hessian regularization for image annotation. IEEE Trans. Syst. Man Cybern. Part B. Cybern. 22(7), 2676–2687 (2013)
Tao, D., Wang, X., Bian, W.: Grassmannian regularized structured multi-view embedding for image classification. IEEE Trans. Image Process. (7):2646–2660
Tuytelaars, Tinne, Mikolajczyk, Krystian: Local invariant feature detectors: a survey. Found. Trends Comput. Graph. Vis. 3(3), 177–280 (2008)
Urtasun, R., Darrell, T.: Discriminative gaussian process latent variable model for classification. In: ICML’2007, (2007)
Wang, Meng, Hua, Xian-Sheng, Hong, Richang, Tang, Jinhui, Qi, Guo-Jun, Song, Yan: Unified video annotation via multigraph learning. IEEE. Trans. Circuits Syst. Video Technol. 19(5), 733–746 (2009)
Wang, Meng, Ni, Bingbing, Hua, Xian-Sheng, Chua, Tat-Seng: Assistive tagging: a survey of multimedia tagging with human-computer joint exploration. ACM Comput. Surv. (CSUR) 44(4), 25 (2012)
Wang, M., Gao, Y., Lu, K., Rui, Y.: View-based discriminative probabilistic modeling for 3d object retrieval and recognition. IEEE Trans. Image Process. 22(4), 1395–1407 (2013)
Xu, C., Tao, D., Xu, C.: Large-margin multi-view information bottleneck. IEEE Trans. Pattern Anal. Mach. Intell.
Xu, C., Tao, D., Li, Y., Xu, C.: Large-margin multi-view gaussian process for image classification. In: ICIMCS’ 2013, (2013a)
Xu, C., Tao, D., Xu, C.: A survey on multi-view learning. arXiv, preprint arXiv:1304.5634, (2013b)
Yu, J., Liu, D., Tao, D., Seah, HS.: On combining multiple features for cartoon character retrieval and clip synthesis. IEEE Trans. Syst. Man Cybern. Part B Cybern, (2012a)
Yu, J., Tao, D.: Modern Machine Learning Techniques and Their Applications in Cartoon Animation Research, vol. 1. Wiley, New York (2013)
Yu, J., Tao, D., Rui, Y., Cheng, J.: Pairwise constraints based multiview features fusion for scene classification. Pattern Recogn. (2012b)
Jun, Yu., Wang, Meng, Tao, Dacheng: Semi-supervised multiview distance metric learning for cartoon synthesis. IEEE Trans. Image Process. 1, 4636–4648 (2012c)
Jun, Y., Rui, Y., Chen, B.: Exploiting click constraints and multi-view features for image re-ranking. IEEE Trans, Multimedia (2013)
Zha, Z.-J., Hua, X.-S., Mei, T., Wang, J., Qi, G.-J., Wang, Z.: Joint multi-label multi-instance learning for image classification. In: CVPR’2008, (2008)
Zha, Z.-J., Yang, L., Mei, T., Wang, M., Wang, Z.: Visual query suggestion. In: MM’2009, (2009)
Zha, Z.-J., Yang, L., Mei, T., Wang, M., Wang, Z., Chua, T.-S., Hua, X.-S.: Visual query suggestion: Towards capturing user intent in internet image search. ACM Trans. Multimedia Comput. Commun. Appl. 6(3), 13 (2010)
Zha, Zheng-Jun, Wang, Meng, Zheng, Yan-Tao, Yang, Yi, Hong, Richang, Chua, Tat-Seng: Interactive video indexing with statistical active learning. IEEE Trans. Multimedia 14(1), 17–27 (2012)
Acknowledgments
The work was supported in part by ARC FT130101457, NBRPC 2011CB302400, NSFC 61121002, 61375026, and JCYJ 20120614152136201.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Xu, C., Tao, D., Li, Y. et al. Large-margin multi-view Gaussian process. Multimedia Systems 21, 147–157 (2015). https://doi.org/10.1007/s00530-014-0389-6
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00530-014-0389-6