At decision module, different match score are met from different modalities and then decision would be taken based on that match score. Although at this level fusion is easy but less powerful due to less availability of adequate information .
2.2 Performance metrics for multimodal biometric system
In password based authentication the perfect match between two alphanumeric strings is necessary to identify or verify persons. In biometric authentication system there should be a perfect match of feature set between a claimed identity and a stored identity. The following terms are uses as performance metrics for multimodal biometric systems:
a) False Match Rate (FMR): It is the probability of a system that incorrectly matches the given input pattern to a non-match stored template in the database. It measures the percentage of invalid inputs which are incorrectly accepted by the system .
b) False Non-Match Rate (FNMR): It is the probability that the system fails to detect a perfect match between the given input pattern and a matched template store in the database. It calculates the percentage of valid inputs which are incorrectly rejected by the system.
c) Receiver Operating Characteristicor Relative Operating Characteristic (ROC): The ROC plot is a curve between the FAR and the FRR. The decision of rejection or acceptance of an individual is taken by comparing the score of the system to a threshold value (called the decision threshold). The values of FRR and FAR are dependent on that threshold value which is to be chosen so as to reduce the global errors of the system .
d) Equal Error Rate or Crossover Error Rate (EER or CER): The curve at which both FAR and FRR errors are equal.
e) Failure To Enrol Rate (FTE): The rate at which attempts to create a template from given input is unsuccessful to enrol. This is most commonly caused by low quality inputs.
f) Failure To Capture rate (FTC): It is the probability of the systems fails to detect a biometric input even when presented correctly.
Figure 6 represent ROC curve and a relationship between FAR (False Accept Rate), FRR (False Reject Rate) and Crossover Error Rate (CER).
Number of person Genuine distribution
0 0.3 η 0.6 0.9
Fig 7: Genuine – Imposter distribution curve
As shown above figure 7 an impostor score that will exceed the threshold η results in a false accept (or, a false match), while a genuine score that will fall below the threshold η results in a false reject (or, a false non-match). The Genuine Accept Rate (GAR) is the fraction of genuine scores exceeding the threshold η value. Therefore,
GAR = 1 – FRR (Fast Reject Rate)
When there are a large number of genuine and impostor scores is available, one could estimate the probability density functions of the two sets of scores in order to analytically derive the FAR (False Accept Rate) and FRR (False Reject Rate) [7,8]. Let p(s|genuine) and p(s|impostor) both represent the probability density functions of the score s under the genuine and impostor conditions, respectively. Then for a particular threshold, η,
FAR(η) = ʃ p(s|impostor)ds,
FRR(η) = ʃ p(s|genuine)ds.
If the match score represents dissimilarity value, then FAR (η) and FRR (η) may be expressed as follows:
FAR(η) = ʃ p(s|impostor)ds,
FRR(η) = ʃ p(s|genuine)ds.
In the case of identification process, the input Feature set is compared against all templates stored in the database in order to determine the best match (i.e. the top match) .
2.3 Existing Technologies of Multimodal Biometrics
Multimodal biometrics uses more than one biometric trait for authentication purpose.
Already some existing multimodal technologies described below table:
Table 1: Existing Multi-Modal Technologies.
Face + Finger veins 
Client specific linear Discriminate analysis(CSLDA)
Face + Speech 
UYVY. AVI 640 x 480, 15.00 fps
Gaussian mixture modal (GMM)
Lip Movement + Gestures 
Faces are Recorded using web camera
Artificial Neural Network (ANN)
Face + Ear 
Principal Component Analysis(PCA)
Face + Ear 
UND-F and FRGC V2
L3DF, Iterative closet point
Face + Ear 
MD I: Yale B and USTB. MD II : AR and USTB
Sparse representation Based classification (SRC), Robust Sparse Coding (RSC)
multi-level ellipse detector combined with a SVM verifier
Face + Palmprint 
AR, PolyU Database
Multimodal biometric systems address some of the problems present in unimodal systems. By combining more than one sources of information, these systems increase population coverage, determination of spoofing, improve matching performance and facilitate indexing. Different type of fusion levels possible in multimodal biometric systems. Fusion at the match score level is the better than others and most popular due to the easy to access and confidence on matching scores. Performance metrics are use to measure the performance or multimodal system in terms of FAR, FRR and ROC curve. Multimodal biometric is spreading for the authentication purpose to maintain the interests of people regarding the security as strong as possible.
REFRENCES: Anil K. Jain, Arun A. Ross, “An Introduction to Biometric Recognition”, IEEE transaction on circuits and systems for video technology, vol.14, no.1, January 2004.
Anil K. Jain Michigan State University, Arun A. Ross West Virginia University, “Multimodal biometrics: An Overview” Appeared in Proc. of 12th European Signal Processing Conference (EUSIPCO), (Vienna, Austria), pp. 1221-1224, September 2004.
Anil K. Jain, Arun A. Ross, “Human Recognition Using Biometrics: An Overview” Appeared in Annals of Telecommunications, Vol. 62, No. 1/2, pp. 11-35, Jan/Feb 2007.
Gandhimathi Amirthalingam, Radhamani, Research Scholar, ”A Multimodal Approach for Face and Ear Biometric System”, IJCSI International Journal of Computer Science Issues, Vol. 10, Issue 5, No 2, September 2013.
Prof. Vijay M. Mane Department of Electronics Engineering, Vishwakarma Institute of Technology, Pune (India), “Review of Multimodal Biometrics: Applications, challenges and Research Areas”.
Muhammad Imran Razzak1, Rubiyah Yusof and Marzuki Khalid,”Multimodal face and finger veins biometric authentication”, Scientific Research and Essays, Vol.5(17), pp. 2529-2534, 2010.
Mohamed Soltane, Noureddine Doghmane, Noureddine Guersi, “Face and Speech Based Multi-Modal Biometric Authentication”, International Journal of Advanced Science and Technology, Vol. 21(8), pp. 41-46, 2010.
A.A. Darwish, R. Abd Elghafar and A. Fawzi Ali, “Multimodal Face and Ear Images”, Journal of Computer Science, Vol. 5 (5), pp. 374-379, 2009.
S.M.S. Islam, R. Davies, M. Bennamoun, R.A. Owens and A.S. Mian, “Multibiometric human recognition using 3D ear and face features”, Pattern Recognition, Vol. 46, No. 3,pp. 613-627, 2013.
Zengxi Huang, Yiguang Liu, Chunguang Li, Menglong Yang and Liping Chen, “A robust face and ear based multimodal biometric system using sparse representation”, Pattern Recognition, Vol. 46, No. 8, pp.2156–2168, 2013.
Xu Xiaona, Pan Xiuqin, Zhao Yue, Pu Qiumei, “Research on Kernel-Based Feature Fusion Algorithm in Multimodal Recognition”, IEEE CS International Conference on Information Technology and Computer Science, pp.3-6, 2009.
Mohammad H. Mahoor , Steven Cadavid, and Mohamed Abdel-Mottaleb, “Multi-modal Ear and Face Modeling and Recognition”, Proc. IEEE 16th International Conference on Image Processing, pp. 4137-4140, 2009.
M. Kawulok, J. Szymanek, “Precise multi-level face detector for advanced analysis of facial images”, IET Image Process.,Vol. 6, Iss. 2, pp. 95–103l, 2012.
B. Dorizzi, “Biometrics at the frontiers, assessing the impact on Society Technical impact of Biometrics”, Background paper for the Institute of Prospective Technological Studies, DG JRC – Sevilla, European Commission, January 2005.
Mohamed Soltane and Mimen Bakhti, “Multi-Modal Biometric Authentications: Concept Issues and Applications Strategies”, International Journal of Advanced Science and Technology Vol. 48, November, 2012.
S. Prabhakar, S. Pankanti, A. K. Jain, “Biometric Recognition: Security and Privacy Concerns”, IEEE Security & Privacy Magazine, vol. 1, no. 2, pp. 33-42, April 2003.