access icon free Strategy of active learning support vector machine for image retrieval

This study proposes a new method for content-based image retrieval by finding an optimal classifier. The optimal classifier is achieved by a new active learning support vector machine (SVM) which combines the model selection with the active learning. The unlabelled samples close to the boundary of the SVM classifier are selected based on the feature similarity for the active learning, and the adaptive regularisation is used to select the optimal model. The combination of model selection with active learning accelerates the convergence of the classifier. The new method can improve the image retrieval accuracy and reduce the time consumption. The experimental results show that the proposed method has a better performance with fewer samples and less time consumption for image retrieval.

Inspec keywords: image retrieval; content-based retrieval; support vector machines; learning (artificial intelligence); optimisation; image classification

Other keywords: model selection; adaptive regularisation; content based image retrieval; optimal classifier; active learning support vector machine; unlabelled sample

Subjects: Optimisation techniques; Information retrieval techniques; Optimisation techniques; Computer vision and image processing techniques; Knowledge engineering techniques; Image recognition

References

    1. 1)
    2. 2)
    3. 3)
    4. 4)
      • 1. Tong, S., Chang, E.: ‘Support vector machine active learning for image retrieval’. Proc. Ninth ACM Int. Conf. on Multimedia, Ottawa, Canada, October 2001, pp. 107118.
    5. 5)
      • 2. Rubens, N., Kaplan, D., Sugiyama, M.: ‘Active learning in recommender systems’, in Kantor, P.B., Ricci, F., Rokach, L., Shapira, B. (Eds.): ‘Recommender systems handbook’ (Springer, 2011), pp. 735767.
    6. 6)
      • 20. Huiskes, M., Lew, M.: ‘Performance evaluation of relevance feedback method’. Proc. 2008 Int. Conf. on Content-Based Image and Video Retrieval (CIVR 08), ACM, New York, July 2008, pp. 239248.
    7. 7)
      • 7. Hastie, T., Rosset, S., Tibshirani, R., et al: ‘The entire regularization path for the support vector machine’, J. Mach. Learn. Res., 2004, 5, (2004), pp. 13911415.
    8. 8)
    9. 9)
      • 9. Steven, C., Jin, R., Zhu, J., et al: ‘Semi-supervised SVM batch mode active learning and its applications to image retrieval’, ACM Trans. Inf. Syst., 2009, 27, (3), pp. 129.
    10. 10)
      • 10. Hu, L., Lu, S., Wang, X.: ‘A new and informative active learning approach for support vector machine’, Inf. Sci., 2013, 244, (2013), pp. 142160.
    11. 11)
      • 6. Vapnik, V.: ‘Statistic learning theory’ (Wiley-Interscience, USA, 1998).
    12. 12)
    13. 13)
    14. 14)
    15. 15)
      • 4. Sugiyama, M., Rubens, N.: ‘Active learning with model selection in linear regression’. Proc. SIAM Int. Conf. on Date Mining (SDM), Atlanta, GA, USA, April 2008, pp. 518529.
    16. 16)
      • 17. Griffin, G., Holub, A., Perona, P.: ‘Caltech-256 object category dataset’. Technical Report 7694, California Institute Technology, 2007.
    17. 17)
    18. 18)
    19. 19)
    20. 20)
      • 15. Nello, C., John, Sh.: ‘An introduction to support vector machine and other kernel-based learning methods’ (Chinese Language Edition Published by Publishing House of Electronics Industry, 2004).
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-cvi.2015.0101
Loading

Related content

content/journals/10.1049/iet-cvi.2015.0101
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading