Abstract
In multimedia annotation, labeling a large amount of training data by human is both time-consuming and tedious. Therefore, to automate this process, a number of methods that leverage unlabeled training data have been proposed. Normally, a given multimedia sample is associated with multiple labels, which may have inherent correlations in real world. Classical multimedia annotation algorithms address this problem by decomposing the multi-label learning into multiple independent single-label problems, which ignores the correlations between different labels. In this paper, we combine label correlation mining and semi-supervised feature selection into a single framework. We evaluate performance of the proposed algorithm of multimedia annotation using MIML, MIRFLICKR and NUS-WIDE datasets. Mean average precision (MAP), MicroAUC and MacroAUC are used as evaluation metrics. Experimental results on the multimedia annotation task demonstrate that our method outperforms the state-of-the-art algorithms for its capability of mining label correlations and exploiting both labeled and unlabeled training data.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Yang, S.H., Hu, B.G.: Feature selection by nonparametric bayes error minimization. In: Proc. PAKDD, pp. 417–428 (2008)
Ma, Z., Nie, F., Yang, Y., Uijlings, J.R.R., Sebe, N., Hauptmann, A.G.: Discriminating joint feature analysis for multimedia data understanding. IEEE Trans. Multimedia 14(6), 1662–1672 (2012)
Nie, F., Henghuang, C.X., Ding, C.: Efficient and robust feature selection via joint l21-norms minimization. In: Proc. NIPS, pp. 759–768 (2007)
Wang, D., Yang, L., Fu, Z., Xia, J.: Prediction of thermophilic protein with pseudo amino acid composition: An approach from combined feature selection and reduction. Protein and Peptide Letters 18(7), 684–689 (2011)
Zhou, Z.H., Zhang, M.L.: Semi-supervised feature selection via spectral analysis. In: Proc. SIAM Int. Conf. Data Mining (2007)
Richard, D., Hart, P.E., Stork, D.G.: Pattern Classification. Wiley-Interscience, New York (2001)
Yang, Y., Wu, F., Nie, F., Shen, H.T., Zhuang, Y., Hauptmann, A.G.: Web and personal image annotation by mining label correlation with relaxed visual graph embedding. IEEE Trans. Image Process. 21(3), 1339–1351 (2012)
Cohen, I., Cozman, F.G., Sebe, N., Cirelo, M.C., Huang, T.S.: Semisupervised learning of classifiers: Theory, algorithms, and their application to human-computer interaction. In: IEEE Trans. PAMI, pp. 1553–1566 (2004)
Zhao, X., Li, X., Pang, C., Wang, S.: Human action recognition based on semi-supervised discriminant analysis with global constraint. Neurocomputing 105, 45–50 (2013)
Wang, S., Ma, Z., Yang, Y., Li, X., Pang, C., Hauptmann, A.: Semi-supervised multiple feature analysis for action recognition. IEEE Trans. Multimedia (2013)
Ji, S., Tang, L., Yu, S., Ye, J.: A shared-subspace learning framework for multi-label classification. ACM Trans. Knowle. Disco. Data 2(1), 8(1)–8(29) (2010)
Ma, Z., Nie, F., Yang, Y., Uijlings, J.R.R., Sebe, N.: Web image annotation via subspace-sparsity collaborated feature selection. IEEE Trans. Multimedia 14(4), 1021–1030 (2012)
Nie, F., Huang, H., Cai, X., Ding, C.: Efficient and robust feature selection via joint l21-norms minimization. In: Proc. NIPS, pp. 1813–1821 (2010)
Yang, Y., Shen, H., Ma, Z., Huang, Z., Zhou, X.: L21-norm regularization discriminative feature selection for unsupervised learning. In: Proc. IJCAI (July 2011)
Zhao, Z., Liu, H.: Spectral feature selection for supervised and unsupervised learning. In: Proc. ICML, pp. 1151–1157 (2007)
Zhou, Z.H., Zhang, M.L.: Multi-instance multi-label learning with application to scene classification. In: Proc. NIPS, pp. 1609–1616 (2006)
Huiskes, M.J., Lew, M.S.: The mir flickr retrieval evaluation. In: Proc. MIR, pp. 39–43 (2008)
Chua, T.S., Tang, J., Hong, R., Li, H., Luo, Z., Zheng, Y.: Nus-wide: A real-world web image database from national university of singapore. In: Proc. CIVR (2009)
Chang, C.C., Lin, C.J.: LIBSVM: A library for support vector machines. ACM Transactions on Intelligent Systems and Technology 2, 27:1–27:27 (2011), http://www.csie.ntu.edu.tw/~cjlin/libsvm
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Chang, X., Shen, H., Wang, S., Liu, J., Li, X. (2014). Semi-supervised Feature Analysis for Multimedia Annotation by Mining Label Correlation. In: Tseng, V.S., Ho, T.B., Zhou, ZH., Chen, A.L.P., Kao, HY. (eds) Advances in Knowledge Discovery and Data Mining. PAKDD 2014. Lecture Notes in Computer Science(), vol 8444. Springer, Cham. https://doi.org/10.1007/978-3-319-06605-9_7
Download citation
DOI: https://doi.org/10.1007/978-3-319-06605-9_7
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-06604-2
Online ISBN: 978-3-319-06605-9
eBook Packages: Computer ScienceComputer Science (R0)