Algorithms of segmentation of speech signal on the correlated noise background
Algorithms of segmentation based on formant and antiformant assessments are considered in the article. An algorithm for speech segmentation using momentary functions of third and fourth order is obtained. It is proposed to use digital filtering based on the sliding-average autoregressive model to su...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PC Technology Center
2018-04-01
|
| Series: | ScienceRise |
| Subjects: | |
| Online Access: | http://journals.uran.ua/sciencerise/article/view/129703 |
| Summary: | Algorithms of segmentation based on formant and antiformant assessments are considered in the article. An algorithm for speech segmentation using momentary functions of third and fourth order is obtained. It is proposed to use digital filtering based on the sliding-average autoregressive model to suppress correlated noise. Estimates of variance in estimating the time boundaries of words for a number of proposed speech segmentation algorithms are obtained |
|---|---|
| ISSN: | 2313-6286 2313-8416 |