Learning a Discriminative Projection and Representation for Image Classification

  • Zuofeng Zhong
  • Jiajun WenEmail author
  • Can Gao
  • Jie Zhou
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 849)


Image classification is a challenging issue in pattern recognition due to complex interior structure for data. Meanwhile, the high-dimension data leads to heavy computational burden. To overcome these shortcomings, in this paper, we learn a discriminative projection and representation in a unified framework for image classification task. This method seeks a discriminative representation in a low-dimension space for an image, which enhances the classification accuracy and efficiency. Thus, a projection matrix is learnt by a criterion which demands the minimum of within-class residual and maximum of between-class residual in an iterative procedure. Then all samples are projected into a low-dimension space, and obtain the discriminative representation via L2 regularization. The experimental results demonstrate that the proposed method achieves better classification performances, compared with state-of-the-art sparse representation methods.


Image classification Discriminative projection Discriminative representation 



This work was supported in part by the Natural Science Foundation of China under Grant 61703283, 61773328, 61672358, 61703169, 61573248, in part by the research grant of the Hong Kong Polytechnic University (Project Code: G-UA2B), in part by the China Postdoctoral Science Foundation under Project 2016M590812, Project 2017T100645 and Project 2017M612736, in part by the Guangdong Natural Science Foundation under Project 2017A030310067, Project with the title Rough Sets-Based Knowledge Discovery for Hybrid Labeled Data and Project with the title The Study on Knowledge Discovery and Uncertain Reasoning in Multi-Valued Decisions.


  1. 1.
    Zhang Z., Xu Y., Shao L., Yang J.: Discriminative block-diagonal representational learning for image recognition. IEEE Transactions on Neural Networks and Learning systems, (2017)MathSciNetCrossRefGoogle Scholar
  2. 2.
    Wright, J., Yang, A.Y., Ganesh, A., Sastry, S.S., Ma, Y.: Robust face recognition via sparse representation. IEEE Trans. Pattern Anal. Mach. Intell. 31(2), 210–227 (2009)CrossRefGoogle Scholar
  3. 3.
    Lai, Z., Xu, Y., Jin, Z., Zhang, D.: Human gait recognition via sparse discriminant projection learning. IEEE Trans. Circuits Syst. Video Technol. 24(10), 1651–1662 (2014)CrossRefGoogle Scholar
  4. 4.
    Wen, J., Lai, Z., Zhan, Y., Cui, J.: The L2,1-norm-based unsupervised optimal feature selection with applications to action recognition. Pattern Recogn. 60, 515–530 (2016)CrossRefGoogle Scholar
  5. 5.
    Zhang L., Yang M., Feng X.: Sparse representation or collaborative representation: Which helps face recognition? In: Proceedings of the International Conference on Computer Vision (ICCV), pp. 471–478 (2011)Google Scholar
  6. 6.
    Yang, J., Chu, D., Zhang, L., Xu, Y., Yang, J.Y.: Sparse representation classifier steered discriminative projection with applications to face recognition. IEEE Trans. Neural Netw. Learn. Syst. 24(7), 1023–1035 (2013)CrossRefGoogle Scholar
  7. 7.
    Belhumeur, P.N., Hespanha, J.P., Kriegman, D.: Eigenfaces vs. fisherfaces: recognition using class specific linear projection. IEEE Trans. Pattern Anal. Mach. Intell. 19(7), 711–720 (1997)CrossRefGoogle Scholar
  8. 8.
    Kim, S.J., Koh, K., Lustig, M., Boyd, S., Gorinevsky, D.: An interior-point method for large-scale l 1-regularized least squares. IEEE J. Sel. Top. Sig. Process. 1(4), 606–617 (2007)CrossRefGoogle Scholar
  9. 9.
    Yang A.Y., Ganesh A., Sastry S.S., Ma, Y.: Fast 1-minimization algorithms and an application in robust face recognition: a review. In: Proceedings of IEEE International Conference on Image Processing, pp. 1849–1852 (2010)Google Scholar
  10. 10.
    Yang, A.Y., Sastry, S.S., Balasubramanian, A.G., Sastry, S.S., Ma, Y.: Fast 1-minimization algorithms for robust face recognition. IEEE Trans. Image Process. 22(8), 3234–3246 (2013)CrossRefGoogle Scholar
  11. 11.
    Beck, A., Teboulle, M.: A fast iterative shrinkage-thresholding algorithm for linear inverse problems. SIAM J. Image Sci. 2(1), 183–202 (2009)MathSciNetCrossRefGoogle Scholar
  12. 12.
    Gross R., Matthews I., Cohn J., Kanade T., Baker S.: Multi-PIE. In: Proceedings of IEEE International Conference on Automatic Face Gesture Recognition, pp. 1–8 (2008)Google Scholar
  13. 13.
    Nene S.A., Nayar S.K., Murase, H.: Columbia object image library (COIL-100). Technical Report CUCS-005-96 (1996)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Zuofeng Zhong
    • 1
  • Jiajun Wen
    • 2
    • 3
    • 4
    Email author
  • Can Gao
    • 2
    • 3
  • Jie Zhou
    • 2
    • 3
  1. 1.Harbin Institute of TechnologyShenzhen Graduate SchoolShenzhenChina
  2. 2.College of Computer Science and Software EngineeringShenzhen UniversityShenzhenChina
  3. 3.Institute of Textiles and ClothingHong Kong Polytechnic UniversityKowloonHong Kong
  4. 4.The Hong Kong Polytechnic University Shenzhen Research InstituteShenzhenChina

Personalised recommendations