Magnolithe GmbH

On Some Features of Log-Spectral Distortion as Speech Quality Measure


<- Back to Automation, Software Development & Engineering Journal Vol. 1

Cite me:

Arkadiy Prodeus (2016). On Some Features of Log-Spectral Distortion as Speech Quality Measure, Automation, Software Development & Engineering Journal, DOI 10.2415/asdej.48.4.664

Authors: Arkadiy Prodeus

ABSTRACT. Evaluation of speech quality for signals passed through low-pass filter showed that log-spectral distortion (LSD) does not decrease monotonically when filter bandwidth is broadening. Experiments have shown that increasing of the analysis time can not compensate for this deficiency. It was suggested that a possible reason for the observed phenomenon is increased sensitivity of LSD to spectral characteristics of speech signal. Analytical and experimental studies have shown the validity of this assumption. To eliminate this drawback, it was proposed averaging of LSD values over set of speakers.

Keywords: speech signal, speech quality, objective quality measure, log-spectral distortion, signal spectrum

References

[1] S. Quackenbush, T. Barnwell, and M. Clements, Objective Measures of Speech Quality. Englewood Cliffs, NJ: Prentice-Hall, 1988.

[2] A. Prodeus and N. Bogdanova, Objective quality evaluation of speech band-limited signals / Prodeus A., Electronics and Communications, 2014, 19 (6), pp. 58-65.

[3] A. Prodeus, K. Zamsha, B. Lozynskiy, Y. Mytiay, and E. Stepanovskaya, Ob’ektivnoe i sub’ektivnoe otsenivanie kachestva rechevyih signalov s ogranichennoy polosoy chastot [Objective and subjective assessment of bandlimited signals speech quality], Electronics and Communications, 2016, 21 (1), pp. 18-26 (in Russian).

[4] A. Prodeus, Reducing Sensitivity of Segmental Signal-to-Noise Ratio Estimator to Time-Alignment Error, International Journal of Electrical and Electronic Science, 2015, 2 (2), pp. 31-36.

[5] A. Prodeus, V. Didkovskiy, and M. Didkovska, Komp”yuterna obrobka akustychnykh syhnaliv. Navchal’nyy posibnyk [Computer processing of acoustic signals. Tutorial]. 2010, Imex-LTD, Kyiv, (in Ukrainian).

[6] F. Nordin, T. Eriksson, A speech spectrum distortion measure with interframe memory, Acoustics, Speech, and Signal Processing, 2001. Proceedings, DOI: 10.1109/ICASSP.2001.941015

Creative Commons Licence
Automation, Software Develoment & Engineering JournalMagnolithe GmbH is licensed under a Creative Commons Attribution 4.0 International License.
Based on a work at www.asdej.xyz.

Search

We assign DOI with

seo4u.link DOI registry