In this paper we proposed new technique for human identification using fusion of both face and speech which can substantially improve the rate of recognition as compared to the single biometric identification for security system development. The proposed system uses principal component analysis (PCA) as feature extraction techniques which calculate the Eigen vectors and Eigen values. These feature vectors are compared using the similarity measure algorithm like Mahalanobis Distances for the

more »

... sion making. The Mel-Frequency cestrum coefficients (MFCC) feature extraction techniques are used for speech recognition in our project. Cross correlation coefficients are considered as primary features. The Hidden Markov Model (HMM) is used to calculate the like hoods in the MFCC extracted features to make the decision about the spoken wards.