| Title |
Comparative analysis of various transformation techniques for voiceless consonants modeling / |
| Authors |
Korvel, Gražina ; Kostek, Bozena ; Kurasova, Olga |
| DOI |
10.15837/ijccc.2018.5.3310 |
| Full Text |
|
| Is Part of |
International journal of computers, communications and control.. Bihor : Agora University Editing House. 2018, vol. 13, no 5, p. 853-864.. ISSN 1841-9836. eISSN 1841-9844 |
| Keywords [eng] |
Discrete Fourier Transform (DFT) ; Discrete Cosine Transform (DCT) ; Discrete Walsh Hadamard Transform (DWHT) ; cepstrum coecients. |
| Abstract [eng] |
In this paper, we propose to use features based on Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT) and Discrete Walsh Hadamard Trans-form (DWHT) for consonant phonemes identification. Feature extraction is to be performed on a speech signal divided into short-time segments. These features are mean and derivative values of cepstrum coefficients. The kNN and Naive Bayes methods are used for phoneme classification. An experiment shows that DFT and DCT give better classification accuracy than DWHT. The result of DFT was not significantly different from DCT, but it was for DWHT. The same tendency was revealed for DCT. The difference between results obtained by DCT and DWHT is significant. |
| Published |
Bihor : Agora University Editing House |
| Type |
Journal article |
| Language |
English |
| Publication date |
2018 |
