Robustness of Score Normalization in Multibiometric Systems
This paper presents an evaluation of normalization techniques of matching scores on the recognition performance of a multibiometric system. We present two score normalization techniques, namely modified-linear-tanh-linear (MLTL) and four-segments-double-sigmoid (FSDS) that are found to be robust in achieving the recognition performance to the optimum value. The techniques are tested in fusion of the two face recognition methods Fisherface and A-LBP on the dataset of uncontrolled environments. In particular, AT & T (ORL) face dataset is used in this experiment. The performance of the MLTL and FSDS score normalization techniques are compared with the existing normalization techniques, for instance min-max, tanh and linear-tanh-linear (LTL). The proposed normalization techniques show the significant improvement in the recognition performance of the multibiometric system over the known techniques.
KeywordsFace recognition Multibiometric Normalization Identification
The authors acknowledge the Institute of Engineering and Technology (IET), Lucknow, Uttar Pradesh Technical University (UPTU), Lucknow for their partial financial support to carry out this research under the Technical Education Quality Improvement Programme (TEQIP-II) grant.
- 8.Shyam, R., Singh, Y.N.: A taxonomy of 2D and 3D face recognition methods. In: Proceedings of 1st International Conference on Signal Processing and Integrated Networks (SPIN 2014), pp. 749–754. IEEE, February 2014Google Scholar
- 9.Shyam, R., Singh, Y.N.: Evaluation of eigenfaces and fisherfaces using bray curtis dissimilarity metric. In: Proceedings of 9th IEEE International Conference on Industrial and Information Systems (ICIIS 2014), pp. 1–6. IEEE, Gwalior, December 2014Google Scholar
- 10.Shyam, R., Singh, Y.N.: Face recognition using augmented local binary patterns and bray curtis dissimilarity metric. In: Proceedings of 2nd International Conference on Signal Processing and Integrated Networks (SPIN 2015), pp. 779–784. IEEE, Noida, February 2015Google Scholar
- 11.Shyam, R., Singh, Y.N.: Analysis of local descriptors for human face recognition. In: Smart Innovation, Systems and Technologies, vol. 43, pp. 263–269. Springer, October 2015Google Scholar
- 12.Shyam, R., Singh, Y.N.: Automatic face recognition in digital world. Adv. Comput. Sci. Inf. Technol. (ACSIT) 2(1), 64–70 (2015)Google Scholar
- 13.Shyam, R., Singh, Y.N.: Recognizing individuals from unconstrained facial images. Adv. Intell. Syst. Comput. Ser. Springer 384, 383–392 (2015)Google Scholar
- 16.Samaria, F., Harter, A.: Parameterisation of a stochastic model for human face identification. In: Proceedings of 2nd IEEE Workshop on Applications of Computer Vision, Sarasota, FL, December 1994Google Scholar