Multimodal biometric scheme for human authentication technique based on voice and face recognition fusion

摘要

In this paper, an effective multimodal biometric identification approach for human authentication tool based on face and voice recognition fusion is proposed. Cepstral coefficients and statistical coefficients are employed to extract features of voice recognition and these two coefficients are compared. Face recognition features are extracted utilizing different extraction techniques, Eigenface and Principle Component Analysis (PCA) and the results are compared. Voice and face identification modality are performed using different three classifiers, Gaussian Mixture Model (GMM), Artificial Neural Network (ANN), and Support Vector Machine (SVM). The combination of biometrics systems, voice and face, into a single multimodal biometric system is performed using features fusion and scores fusion. The computer simulation experiments reveal that better results are given in case of utilizing for voice recognition the cepstral coefficients and statistical coefficients and in case of face, Eigenface and SVM experiment gives better results for face recognition. Also, in the proposed multimodal biometrics system the scores fusion performs better than other scenarios.