Voice Biometric Identity Authentication System Based on Android Smart Phone
Authors:
- 1 December 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1440-1444
Abstract
This thesis studies and implements voice biometric authentication system on Android smart phone platform. Firstly, the characteristics of voice database and voice recognition and authentication system flow are introduced. Then, a random shuffling algorithm is used to generate a voice cipher. After this, the voice feature extraction algorithm based on MFCC is introduced. In this paper, GMM is introduced to train the voiceprint feature model of voice segment data effectively. When number of training samples is 5, the success rate of system authentication is between 89% and 96% and time required to authenticate a voice sample is between 210ms~320ms, which demonstrate the high accuracy and good real-time.Keywords
This publication has 6 references indexed in Scilit:
- Biometric-Enabled Authentication Machines: A Survey of Open-Set Real-World ApplicationsIEEE Transactions on Human-Machine Systems, 2015
- A unique approach in text independent speaker recognition using MFCC feature sets and probabilistic neural network2015
- Telephone Speech Endpoint Detection using Mean-Delta FeatureCybernetics and Information Technologies, 2014
- Silence Removal and Endpoint Detection of Speech Signal for Text Independent Speaker IdentificationInternational Journal of Image, Graphics and Signal Processing, 2014
- Biometric authentication as a service for enterprise identity management deployment: a data protection perspective2011
- Speaker Verification Using Adapted Gaussian Mixture ModelsDigital Signal Processing, 2000