Real time user authentication system for PDA

Personal digital assistants (PDA), a representative example of current and future class of portable devices, contain important personal information of the users, which relate closely to users' privacy and require increased level of protection. This paper presents an automatic speaker verification system for PDA which facilitates reliable and convenient user authentication. We apply adaptive double thresholds endpoint detection (ADED) and feature clipping (FC) to make the system robust to ambient noise. The whole system implementation is simple and compact enough to fit well with PDA's limited memory and computational power. The system runs on Compaq's ipaq3630. Batch test and field running have been taken for our system, the results are promising

[1]  ChiWei Che,et al.  Speaker recognition using HMM with experiments on the yoho database , 1995, EUROSPEECH.

[2]  Ken-ichi Iso,et al.  An automatic speech translation system on PDAs for travel conversation , 2002, Proceedings. Fourth IEEE International Conference on Multimodal Interfaces.

[3]  Deng Yong.  SPEECH RECOGNITION ALGORITHM DESIGN AND IMPLEMENTATION ON PALM PC , 2000 .

[4]  Biing-Hwang Juang,et al.  Fundamentals of speech recognition , 1993, Prentice Hall signal processing series.

[5]  Douglas A. Reynolds,et al.  Robust text-independent speaker identification using Gaussian mixture speaker models , 1995, IEEE Trans. Speech Audio Process..

[6]  Jr. J.P. Campbell,et al.  Speaker recognition: a tutorial , 1997, Proc. IEEE.