ISCA Archive SPECOM 2004
ISCA Archive SPECOM 2004

Speech signal analysis wavelet-transformation and signal processing at the periphery of acoustical system

Vladimir Bondarenko, Vladislav Kotsubinski, Andrew Ponomarev, Dmitriy Velikotski

The report deals with problems of speech signal processing in speech recognition systems being analogue of aural perception system. It’s known from psychoacoustics and neurophysiology of perception that latter possesses high resolution on frequency and time. It’s rather difficult to meet theses conflicting objectives with the help of Fourier analysis. Nowadays various species of Wavelet transformation are widely applied to analysis of signals of different classes. Such transformations are aimed at compromise achievement between frequency and time resolutions during research of signal structure. Moreover Wavelet-transformation intended for computer-based signal processing from the very outset. One could suppose that at the periphery of acoustical system input signal processing similar to Wavelet-transformation is realized ensuring required frequency and time resolution. Such data as transfer constants of internal ear basic membrane; data on logarithmic frequency scale; data on consecutive and synchronous masking can be considered as arguments of the aforesaid assumption. The report contain analysis of all these data demonstrating that sufficiently specific transformation of input signals is implemented at the periphery of acoustical system. This transformation can be placed among classes of Wavelet-transformation. However the named transformation possesses /distinctive features allowing discovering of perceived signal structure peculiarities.


Cite as: Bondarenko, V., Kotsubinski, V., Ponomarev, A., Velikotski, D. (2004) Speech signal analysis wavelet-transformation and signal processing at the periphery of acoustical system. Proc. 9th Conference on Speech and Computer (SPECOM 2004), 181-185

@inproceedings{bondarenko04_specom,
  author={Vladimir Bondarenko and Vladislav Kotsubinski and Andrew Ponomarev and Dmitriy Velikotski},
  title={{Speech signal analysis wavelet-transformation and signal processing at the periphery of acoustical system}},
  year=2004,
  booktitle={Proc. 9th Conference on Speech and Computer (SPECOM 2004)},
  pages={181--185}
}