International Journal of Computer Vision

, Volume 91, Issue 2, pp 216–232 | Cite as

Incremental Linear Discriminant Analysis Using Sufficient Spanning Sets and Its Applications

  • Tae-Kyun KimEmail author
  • Björn Stenger
  • Josef Kittler
  • Roberto Cipolla


This paper presents an incremental learning solution for Linear Discriminant Analysis (LDA) and its applications to object recognition problems. We apply the sufficient spanning set approximation in three steps i.e. update for the total scatter matrix, between-class scatter matrix and the projected data matrix, which leads an online solution which closely agrees with the batch solution in accuracy while significantly reducing the computational complexity. The algorithm yields an efficient solution to incremental LDA even when the number of classes as well as the set size is large. The incremental LDA method has been also shown useful for semi-supervised online learning. Label propagation is done by integrating the incremental LDA into an EM framework. The method has been demonstrated in the task of merging large datasets which were collected during MPEG standardization for face image retrieval, face authentication using the BANCA dataset, and object categorisation using the Caltech101 dataset.


Linear discriminant analysis LDA Incremental learning Online learning Label propagation Semi-supervised learning Face image retrieval Object recognition Object categorisation Face authentication 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Bar-Hillel, A., Hertz, T., Shental, N., & Weinshall, D. (2005). Learning a Mahalanobis metric from equivalence constraints. Journal of Machine Learning Research, 6, 937–965. MathSciNetGoogle Scholar
  2. Belkin, M., & Niyogi, P. (2004). Learning with local and global consistency. In Proc. of NIPS. Google Scholar
  3. Berg, A., Berg, T., & Malik, J. (2005). Shape matching and object recognition using low distortion correspondence. In Proc. of CVPR. Google Scholar
  4. Bouveyron, C., Girard, S., & Schmid, C. (2004). Dimension reduction and classification methods for object recognition in vision. In Proc. of 5th French-Danish workshop on spatial statistics and image analysis in biology, Saint-Pierre de Chartreuse, France, May 2004. Google Scholar
  5. Chan, C.-H., Kittler, J., & Messer, K. (2007). Multi-scale local binary pattern histograms for face recognition. In Proc. of ICB (pp. 809–818). Google Scholar
  6. Cheng, H., Liu, Z., & Yang, J. (2009). Sparsity induced similarity measure for label propagation. In Proc. of ICCV. Google Scholar
  7. Chin, T.-J., & Suter, D. (2006). Incremental kernel PCA for efficient non-linear feature extraction. In Proc. of BMVC. Google Scholar
  8. De la Torre Frade, F., Gross, R., Baker, S., & Kumar, V. (2005). Representational oriented component analysis (ROCA) for face recognition with one sample image per training class. In Proc. of CVPR. Google Scholar
  9. Fei-Fei, L., Fergus, R., & Perona, P. (2004). Learning generative visual models from few training examples. In Proc. of CVPR workshop on GMBV. Google Scholar
  10. Fukunaga, K. (1990). Introduction to statistical pattern recognition. Boston: Academic Press. zbMATHGoogle Scholar
  11. Grauman, K., & Darrell, T. (2005). The pyramid match kernel: discriminative classification with sets of image features. In Proc. of ICCV. Google Scholar
  12. Hall, P., Marshall, D., & Martin, R. (2000). Merging and splitting eigenspace models. IEEE Transactions on Pattern Analysis and Machine Intelligence, 22(9), 1042–1049. CrossRefGoogle Scholar
  13. Hiraoka, K., Hidai, K., Hamahira, M., Mizoguchi, H., Mishima, T., & Yoshizawa, S. (2000). Successive learning of linear discriminant analysis: Sanger-type algorithm. In Proc. of ICPR. Google Scholar
  14. Holub, A. D., Welling, M., & Perona, P. (2005). Exploiting unlabelled data for hybrid object classification. In Proc. of NIPS workshop on inter-class transfer. Google Scholar
  15. Kamei, T., Yamada, A., Kim, T., Kim, H., Hwang, W., & Kee, S. (2002). Advanced face descriptor using Fourier and intensity LDA features. ISO/IEC JTC1/SC29/WG11 M8998, October 2002. Google Scholar
  16. Kim, H.-C., Kim, D., & Bang, S. Y. (2003). Face recognition using LDA mixture model. Pattern Recognition Letters, 24(15), 2815–2821. CrossRefGoogle Scholar
  17. Kim, T.-K., Kim, H., Hwang, W., & Kittler, J. (2005). Component-based LDA face description for image retrieval and MPEG-7 standardisation. Image and Vision Computing, 23, 631–642. CrossRefGoogle Scholar
  18. Kim, T.-K., Kittler, J., & Cipolla, R. (2007). Discriminative learning and recognition of image set classes using canonical correlations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(6), June 2007. Google Scholar
  19. Kim, T.-K., Wong, S.-F., Stenger, B., Kittler, J., & Cipolla, R. (2007). Incremental linear discriminant analysis using sufficient spanning set approximations. In Proc. of CVPR, Minneapolis, MN. Google Scholar
  20. Leibe, B., & Schiele, B. (2003). Analyzing appearance and contour based methods for object categorization. In Proc. of CVPR (pp. 409–415). Google Scholar
  21. Lin, R.-S., Ross, D., Lim, J., & Yang, M.-H. (2005). Adaptive discriminative generative model and its applications. In Proc. of NIPS. Google Scholar
  22. Manjunath, B. S., Salembier, P. P., & Sikora, T. (2002). Introduction to MPEG-7: multimedia content description interface. New York: Wiley. Google Scholar
  23. Messer, K. et al. (2004). Face authentication competition on the BANCA database. In Proc. of ICPR. Google Scholar
  24. Moosmann, F., Triggs, B., & Jurie, F. (2007). Fast discriminative visual codebooks using randomized clustering forests. In Proc. of NIPS. Google Scholar
  25. Mutch, J., & Lowe, D. (2006). Multiclass object recognition with sparse, localized features. In Proc. of CVPR. Google Scholar
  26. Niebles, J. C., Wang, H., & Fei-Fei, L. (2008). Unsupervised learning of human action categories using spatial-temporal words. International Journal of Computer Vision. Google Scholar
  27. Oja, E. (1983). Subspace methods of pattern recognition. Research Studies Press. Google Scholar
  28. Pang, S., Ozawa, S., & Kasabov, N. (2005). Incremental linear discriminant analysis for classification of data streams. IEEE Transactions on System, Man and Cybernetics, 905–914. Google Scholar
  29. Poh, N. et al. (2009). Face video competition. In Proc. of ICPR. Google Scholar
  30. Skocaj, D., & Leonardis, A. (2003). Weighted and robust incremental method for subspace learning. In Proc. of ICCV. Google Scholar
  31. Stenger, B., Woodley, T., Kim, T.-K., Hernandez, C., & Cipolla, R. (2008). AIDIA—adaptive interface for display interaction. In Proc. of BMVC, Leeds, UK. Google Scholar
  32. Tao, X., Ye, J., Li, Q., Janardan, R., & Cherkassky, V. (2004). Efficient kernel discriminant analysis via QR decomposition. In Proc. of NIPS. Google Scholar
  33. Uray, M., Skocaj, D., Roth, P., Bischof, H., & Leonardis, A. (2007). Incremental LDA learning by combining reconstructive and discriminative approaches. In Proc. of BMVC. Google Scholar
  34. Wang, F., & Zhang, C. (2007). Label propagation through linear neighborhoods. In Proc. of ICML. Google Scholar
  35. Wang, X., Han, T. X., & Yan, S. (2009). An HOG-LBP human detector with partial occlusion handling. In Proc. of ICCV, Kyoto. Google Scholar
  36. Winn, J., Criminisi, A., & Minka, T. (2005). Object categorisation by learned universal visual dictionary. In Proc. of ICCV. Google Scholar
  37. Wu, Y., & Huang, S. (2000). View-independent recognition of hand postures. In Proc. of CVPR (pp. 2088–2094). Google Scholar
  38. Yan, J., Zhang, B., Yan, S., Yang, Q., & Li, H. (2004). IMMC: incremental maximum margin criterion. In Proc. of int’l conf. knowledge discovery and data mining. Google Scholar
  39. Ye, J., Li, Q., Xiong, H., Park, H., Janardan, V., & Kumar, V. (2005). IDR/QR: an incremental dimension reduction algorithm via QR decomposition. IEEE Transactions on Knowledge and Data Engineering, 17(9), 1208–1222. CrossRefGoogle Scholar
  40. Yu, H., & Yang, H. (2001). A direct LDA algorithm for high-dimensional data—with application to face recognition. Pattern Recognition, 34(10), 2067–2070. zbMATHCrossRefGoogle Scholar
  41. Zhu, X. (2006). Semi-supervised learning literature survey. Computer Sciences TR 1530, University of Wisconsin-Madison. Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • Tae-Kyun Kim
    • 1
    Email author
  • Björn Stenger
    • 2
  • Josef Kittler
    • 3
  • Roberto Cipolla
    • 4
  1. 1.Sidney Sussex CollegeUniversity of CambridgeCambridgeUK
  2. 2.Toshiba Research Europe LtdCambridgeUK
  3. 3.Centre for Vision, Speech and Signal ProcessingUniversity of SurreyGuildfordUK
  4. 4.Department of EngineeringUniversity of CambridgeCambridgeUK

Personalised recommendations