• Rajni
• Suraj Pal

In today’s informative world, identifying a person’s identity accurately and protecting the security is becoming important. Biometric recognition technology is becoming crucial and is widely being used. In today’s world the most convenient and secured solution for the identification is multimodal biometric identification, which single biometric identification cannot achieve because of the complex identification situations. With multimodal system, accuracy and safety is achieved which may be lacked in single identification system. In this paper, biometric recognition system based on face, iris and fingerprints with CNN are evaluated. Multimodal biometric system based on deep learning algorithm is recommended for identifying humans based on face, iris and fingerprints. The whole structure of the system is based on convolutional neural networks (CNN). To find out the result of the accuracy on recognition system, CNN model is made for unimodal recognition system. Different fusion approaches are applied to carry out the recognition system. Then the CNN model for multimodal biometric system is developed based on two- layer fusion. In this paper Alex-Net and VGG-19 network models are evaluated in experimentation part for extracting iris, fingerprint and face image features as an input to the feature fusion module. Most of the empirical work is conducted using CMU PIE, CASIA and POLY-U datasets. Later based on both the studies it is concluded that multimodal biometric system is more accurate and reliable as compared to unimodal system. Furthermore, improvement to the multimodal biometric systems in terms of multi focal loss function for feature extraction was suggested.

Multimodal Biometrics, Convolutional neural network, face recognition, iris recognition, and two-layer fusion.

"Mechanism to increase the security using multimodal biometric with different approaches", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 9, page no.f89-f101, September-2023, Available :http://www.jetir.org/papers/JETIR2309509.pdf

"Mechanism to increase the security using multimodal biometric with different approaches", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 9, page no. ppf89-f101, September-2023, Available at : http://www.jetir.org/papers/JETIR2309509.pdf