PROPOSITION OF MINIMUM BANDS MULTIRATE NOISE REDUCTION SYSTEM WHICH EXPLOITS PROPERTIES OF THE HUMAN AUDITORY SYSTEM AND ALL-PASS TRANSFORMED FILTER BANK †

This paper introduce the new approach to noise reduction in order to improve the speech intelligibility. The system is proposed with minimum band requirement to approximate psychoacoustic Bark scale and nonuniform filter bank constructed with the use of first order all-pass transformation. Proposed psychoacoustic weighting exploits the well known audible noise reduction rule in case of the 16-bit signal and sampling rate of 8 kHz. Mainly all solutions exploit the masking property of the human ear and calculate the excitation pattern and masking threshold based on Fourier analysis or Wavelet transform. Mentioned Fourier decomposition method provides the frequency resolution much greater than the auditory Bark scale commonly used. While the wavelet transform provides the step-linear approximation of the auditory scale. Then by simple addition of the overload decomposition calculate the psychoacoustic information in the Bark bands. In this paper the multirate system based on filter bank with the quantity of bands equal the number of Barks for assumed sampling frequency is proposed. Solution is mentioned for the 8 kHz sampling frequency in hands-free device, which is a minimum cost noise reduction system considering the number of bands and sampling frequency but still satisfying the intelligibility improvement of the noisy speech. 2 Audible noise suppression rule

[1]  A. Oppenheim,et al.  Computation of spectra with unequal resolution using the fast Fourier transform , 1971 .

[2]  Andrzej Czyzewski,et al.  Noise reduction algorithms employing an intelligent inference engine for multimedia applications , 1998, 1998 IEEE Second Workshop on Multimedia Signal Processing (Cat. No.98EX175).

[3]  James D. Johnston,et al.  Transform coding of audio signals using perceptual noise criteria , 1988, IEEE J. Sel. Areas Commun..

[4]  R. Haddad,et al.  Multiresolution Signal Decomposition: Transforms, Subbands, and Wavelets , 1992 .

[5]  Olivier Cappé,et al.  Elimination of the musical noise phenomenon with the Ephraim and Malah noise suppressor , 1994, IEEE Trans. Speech Audio Process..

[6]  Peter Jax,et al.  A novel psychoacoustically motivated audio enhancement algorithm preserving background noise characteristics , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[7]  Andreas Engelsberg,et al.  Comparison of a discrete wavelet transformation and a nonuniform polyphase filterbank applied to spectral-subtraction speech enhancement , 1998, Signal Process..

[8]  Gerhard Doblinger,et al.  Computationally efficient speech enhancement by spectral minima tracking in subbands , 1995, EUROSPEECH.

[9]  Schuyler Quackenbush,et al.  Objective measures of speech quality , 1995 .

[10]  Nathalie Virag,et al.  Single channel speech enhancement based on masking properties of the human auditory system , 1999, IEEE Trans. Speech Audio Process..

[11]  H. Strube.  Linear prediction on a warped frequency scale , 1980 .

[12]  Truong Q. Nguyen,et al.  Wavelets and filter banks , 1996 .

[13]  Richard M. Schwartz,et al.  Enhancement of speech corrupted by acoustic noise , 1979, ICASSP.

[14]  Henning Puder,et al.  Speech enhancement for mobile telephony based on non-uniformly spaced frequency resolution , 1998, 9th European Signal Processing Conference (EUSIPCO 1998).

[15]  Tim Haulick,et al.  Residual noise suppression using psychoacoustic criteria , 1997, EUROSPEECH.

[16]  S. Biyiksiz,et al.  Multirate digital signal processing , 1985, Proceedings of the IEEE.

[17]  Sofia Ben Jebara,et al.  Reduction of musical noise generated by spectral subtraction by combining wavelet packet transform and wiener filtering , 2000, 2000 10th European Signal Processing Conference.

[18]  Thomas F. Quatieri,et al.  Noise reduction based on spectral change , 1997, Proceedings of 1997 Workshop on Applications of Signal Processing to Audio and Acoustics.

[19]  P. Vaidyanathan.  Multirate Systems And Filter Banks , 1992 .

[20]  Ephraim.  Speech enhancement using a minimum mean square error short-time spectral amplitude estimator , 1984 .

[21]  Julius O. Smith,et al.  Bark and ERB bilinear transforms , 1999, IEEE Trans. Speech Audio Process..

[22]  David Malah,et al.  Speech enhancement using a minimum mean-square error log-spectral amplitude estimator , 1984, IEEE Trans. Acoust. Speech Signal Process..

[23]  Christophe Beaugeant,et al.  Noise reduction using perceptual spectral change , 1999, EUROSPEECH.

[24]  L. Deng.  Ieee Transactions on Speech and Audio Processing, Speech Trajectory Discrimination Using the Minimum Classiication Error Learning , 1997 .

[25]  S. Boll,et al.  Suppression of acoustic noise in speech using spectral subtraction , 1979 .

[26]  John Mourjopoulos,et al.  Improving the intelligibility of noisy speech using an audible noise suppression technique , 1997, EUROSPEECH.

[27]  Anthony G. Constantinides,et al.  Spectral Transformations for Digital Filters , 1970 .