FRACTIONAL FOURIER TRANSFORM COMBINATION WITH MFCC BASED SPEAKER IDENTIFICATION IN CLEAN ENVIRONMENT

The Fractional Fourier Transform FrFT transform order for the proper analysis of multi-component signa ls like speech is still debated. In this paper we have comp ared as well as verified the technique of speaker i dentification using FrFT combined with Mel-frequency cepstral coefficients (MFCC). The FrFT with MFCC shows poor result as compare to MFCC with FFT. In text- independent case the difference in result is 3.41%, while in the te xt-dependent case difference becomes 13.62%. This is because FrFT emphasis only harmonics of speech signal.

[1]  R. Tao,et al.  Detection and parameter estimation of multicomponent LFM signal based on the fractional fourier transform , 2004, Science in China Series F: Information Sciences.

[2]  H. Ezzaidi,et al.  Pitch and MFCC dependent GMM models for speaker identification systems , 2004, Canadian Conference on Electrical and Computer Engineering 2004 (IEEE Cat. No.04CH37513).

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

[4]  Liang He,et al.  Fractional Fourier transform based auditory feature for language identification , 2008, APCCAS 2008 - 2008 IEEE Asia Pacific Conference on Circuits and Systems.

[5]  Tomi Kinnunen,et al.  Real-time speaker identification and verification , 2006, IEEE Transactions on Audio, Speech, and Language Processing.

[6]  B. Atal.  Effectiveness of linear prediction characteristics of the speech wave for automatic speaker identification and verification. , 1974, The Journal of the Acoustical Society of America.

[7]  T. Parks,et al.  Maximum likelihood pitch estimation , 1976, 1977 IEEE Conference on Decision and Control including the 16th Symposium on Adaptive Processes and A Special Symposium on Fuzzy Set Theory and Applications.