Search
Highlighting interlanguage phoneme differences based on similarity matrices and convolutional neural network
Abstract
The goal of this research is to find a way of highlighting the acoustic differences between consonant phonemes of the Polish and Lithuanian languages. For this purpose, similarity matrices are employed based on speech acoustic parameters combined with a convolutional neural network (CNN). In the first experiment, we compare the effectiveness of the similarity matrices applied to discerning acoustic differences between consonant phonemes of the Polish and Lithuanian languages. The similarity matrices built on both an extensive set of parameters and a reduced set after removing high-correlated parameters are used. The results show that higher accuracy is obtained by the similarity matrices without discarding high-correlated parameters. In the second experiment, the averaged accuracies of the similarity matrices obtained are compared with the results provided by spectrograms combined with CNN, as well as the results of the vectors containing acoustic parameters and two baseline classifiers, namely k-nearest neighbors and support vector machine. The performance of the similarity matrix approach demonstrates its superiority over the methods used for comparison.
Citations
-
6
CrossRef
-
0
Web of Science
-
8
Scopus
Authors (3)
-
Grazina Korvel
-
Povilas Treigys dr
Cite as
Full text
- Publication version
- Accepted or Published Version
- License
-
open in new tab
Keywords
Details
- Category:
- Articles
- Type:
- artykuły w czasopismach
- Published in:
-
Journal of the Acoustical Society of America
no. 149,
pages 508 - 523,
ISSN: 0001-4966 - Language:
- English
- Publication year:
- 2021
- Bibliographic description:
- Korvel G., Treigys P., Kostek B.: Highlighting interlanguage phoneme differences based on similarity matrices and convolutional neural network// Journal of the Acoustical Society of America -Vol. 149,iss. 1 (2021), s.508-523
- DOI:
- Digital Object Identifier (open in new tab) 10.1121/10.0003339
- Sources of funding:
-
- Statutory activity/subsidy
- Verified by:
- Gdańsk University of Technology
seen 113 times
Recommended for you
INVESTIGATION OF THE LOMBARD EFFECT BASED ON A MACHINE LEARNING APPROACH
- G. Korvel,
- P. Treigys,
- K. Kąkol
- + 1 authors
Comparison of Lithuanian and Polish Consonant Phonemes Based on Acoustic Analysis – Preliminary Results
- G. Korvel,
- O. Kurasova,
- B. Kostek
A Study of Cross-Linguistic Speech Emotion Recognition Based on 2D Feature Spaces
- G. Tamulevicius,
- G. Korvel,
- A. B. Yayak
- + 3 authors