SPEAKER CHARACTERISTICS FROM A GLOTTAL AIRFLOW MODEL USING ROBUST INVERSE FILTERING

Here the excitation, u(n), is the volume velocity or volume flow rate of air through the glottis, V(z) is a linear filter representing the vocal tract and R(z) is a linear filter representing the lip radiation. The input to R(z) is the volume velocity at the lips while the output, s(n), is the pressure waveform at the microphone. For frequencies below a few kHz, R(z) approximates a differentiator [20] and in this work it is taken to be (1 1 − − z ). Providing the characteristics of the vocal tract do not change significantly during the impulse response of R(z), we can interchange V(z) and R(z) without altering the output signal. The input to V(z) is now u'(n) , the first difference of the glottal airflow waveform:

[1]  B. Yegnanarayana,et al.  Epoch extraction from linear prediction residual for identification of closed glottis interval , 1979 .

[2]  R. Miller Nature of the Vocal Cord Wave , 1956 .

[3]  D. Veeneman,et al.  Automatic glottal inverse filtering from speech and electroglottographic signals , 1985, IEEE Trans. Acoust. Speech Signal Process..

[4]  A. K. Krishnamurthy Two channel (speech and egg) analysis for formant and glottal inverse filtering , 1984, ICASSP.

[5]  A. Gray,et al.  Least squares glottal inverse filtering from the acoustic speech waveform , 1979 .

[6]  H. W. Strube Determination of the instant glottal closure from the speech waveform , 1974 .

[7]  Jr. J. Deller Some notes on closed phase glottal inverse filtering , 1981 .

[8]  Evelyn Abberton,et al.  Laryngographic assessment of normal voice: A tutorial , 1989 .

[9]  John H. L. Hansen,et al.  Discrete-Time Processing of Speech Signals , 1993 .

[10]  Hiroya Fujisaki,et al.  Estimation of voice source and vocal tract parameters based on ARMA analysis and a model for the Glottal source waveform , 1987, ICASSP '87. IEEE International Conference on Acoustics, Speech, and Signal Processing.

[11]  A. Rosenberg Effect of glottal pulse shape on the quality of natural vowels. , 1969, The Journal of the Acoustical Society of America.

[12]  Yves Kamp,et al.  A Frobenius norm approach to glottal closure detection from the speech signal , 1994, IEEE Trans. Speech Audio Process..

[13]  Donald G. Childers,et al.  Glottal area versus glottal volume-velocity , 1977 .

[14]  A. K. Krishnamurthy,et al.  TWO CHANNEL (SPEECH AND EGG) ANALYSIS FOR FORMANT TRACKING AND GLOTTAL INVERSE FILTERING , 1984 .

[15]  A. P. Lobo,et al.  Evaluation of a glottal ARMA model of speech production , 1992, [Proceedings] ICASSP-92: 1992 IEEE International Conference on Acoustics, Speech, and Signal Processing.