Deep Learning based Fusion for a Multi-Biometric Identification Using LSTM

Now a days for security aspect identify a person with his/her biometric traits are very popular as well as applicable also with high accuracy. Many of the applications use unimodal biometric identification but in unimodal biometric identification system accuracy is very low. These combinations of two or more than two biometric traits give high accuracy. Here defines the comparison of single and multibiometric recognition approaches and role of deep learning based fusion in multimodel identification process. In this paper proposed a LSTM (Long Short Term Memory) model for re-recognize a person using Face and Gait Biometric trains with high accuracy. The proposed method gives 99.67% accuracy. The results point out that the biometric acknowledgment system using deep learning is secure, robust, and reliable. The use of this LSTM approach is in such area where high accuracy needed. such as police or defense organization where need high accuracy. We can also use this framework when we authenticate a person by drone.