Abstract
The article considers an algorithm for segmenting a speech signal into sections of silence and voiced and unvoiced segments. The use of the segmentation algorithm will make it possible to assess the quality of speech based on the comparison of two different implementations of the same phoneme separately, not as part of a whole syllable. The algorithm under study is based on the classification of individual signal frames into classes of silence, unvoiced or voiced section and the selection of segments by combining adjacent frames of the same classes. Testing showed the efficiency of the algorithm both on normal (undistorted) speech and on distorted speech of patients undergoing speech rehabilitation. An algorithm was investigated with class definition parameters proposed by the authors of the algorithm. Also, new classification parameters were proposed, selected using methods for optimizing the values of the exposed segment boundaries based on the analysis of a dataset from normal and distorted speech. The use of optimally selected parameters made it possible to reduce the segmentation error by an average of 60%. The applicability of the considered segmentation algorithm for solving the problem of dividing syllables into phonemes is shown.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Kaprin, A.D., Starinskiy, V.V., Shahzadova, A.O.: Malignancies in Russia in 2021 (Morbidity and mortality), 252 p. MNIOI name of P.A. Herzen, Moscow (2022)
Kaprin, A.D., Starinskiy, V.V., Shahzadova, A.O.: The state of oncological care for the population of Russia in 2021, 239 p. MNIOI name of P.A. Herzen, Moscow (2022)
Kostuchenko, E., et al.: The Evaluation Process Automation of Phrase and Word Intelligibility Using Speech Recognition Systems. In: Albert Ali Salah, Alexey Karpov, Rodmonga Potapova, (ed.) Speech and Computer: 21st International Conference, SPECOM 2019, Istanbul, Turkey, August 20–25, 2019, Proceedings, pp. 237–246. Springer International Publishing, Cham (2019). https://doi.org/10.1007/978-3-030-26061-3_25
Novokhrestova D. I., Kostyuchenko E. Yu., Hodashinsky I. A. Algorithm and method of quantitative assessment of the similarity of speech signals // Proceedings of TUSUR University. - 2022. - V. 25, No. 3. - P. 45–51 (In Russian). DOI: https://doi.org/10.21293/1818-0442-2022-25-3-45-51
Kharchenko, S.S., Novokhrestova, D.I., Kostyuchenko, E.: The problem of segmentation into phonemes in assessing the quality of pronunciation of syllables in the framework of speech rehabilitation. Mater. Int. Sci. Conf. Electron. Devices Control Syst. 1–1, 223–226 (2018). (In Russian)
Cai, R.: A modified multi-feature voiced/unvoiced speech classification method. In: 2010 Asia-Pacific Conference on Power Electronics and Design. IEEE, pp. 68–71 (2010)
Atal, B., Rabiner, L.: A pattern recognition approach to voiced-unvoiced-silence classification with applications to speech recognition. IEEE Trans. Acoust. Speech Signal Process. 24(3), 201–212 (1976)
Ijitona, T., Yue, H., Soraghan, J., Lowit, A.: Improved silence-unvoiced-voiced (SUV) segmentation for dysarthric speech signals using linear prediction error variance. In: 2020 5th International Conference on Computer and Communication Systems (ICCCS). IEEE, 2020, pp. 685–690 (2020)
Warule, P., Mishra, S.P., Deb, S.: Significance of voiced and unvoiced speech segments for the detection of common cold. Signal, Image and Video Process. 17(5), 1785–1792 (2023). https://doi.org/10.1007/s11760-022-02389-8
MATLAB R2023a, https://www.mathworks.com/products/matlab.html, Accessed 5 May 2023
Wavesurfer, https://sourceforge.net/projects/wavesurfer/, Accessed 5 May 2023
Lagarias, J., Reeds, J., Wright, M., Wright, P.: Convergence properties of the nelder-mead simplex method in low dimensions. SIAM J. Optim. 9(1), 112–147 (1998)
Byrd, R., Gilbert, J., Nocedal, J.: A trust region method based on interior point techniques for nonlinear programming. Math. Program. 89(1), 149–185 (2000)
Wilcoxon, F.: Individual comparisons by ranking methods. Biometrics Bulletin, № 1 (6), 12, pp. 80–83 (1945)
Acknowledgments
Acknowledgments. This research was funded by the Ministry of Science and Higher Education of the Russian Federation within the framework of scientific projects carried out by teams of research laboratories of educational institutions of higher education subordinate to the Ministry of Science and Higher Education of the Russian Federation, project number FEWM-2020–0042.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Novokhrestova, D., Kostyuchenko, E., Krivoshein, I., Balatskaya, L. (2023). Speech Signal Segmentation into Silence, Unvoiced and Vocalized Sections in Speech Rehabilitation. In: Karpov, A., Samudravijaya, K., Deepak, K.T., Hegde, R.M., Agrawal, S.S., Prasanna, S.R.M. (eds) Speech and Computer. SPECOM 2023. Lecture Notes in Computer Science(), vol 14338. Springer, Cham. https://doi.org/10.1007/978-3-031-48309-7_48
Download citation
DOI: https://doi.org/10.1007/978-3-031-48309-7_48
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-48308-0
Online ISBN: 978-3-031-48309-7
eBook Packages: Computer ScienceComputer Science (R0)