| Autor: |
Mou Hua |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
Applied Mathematics and Nonlinear Sciences, Vol 8, Iss 1, Pp 1377-1384 (2023) |
| Druh dokumentu: |
article |
| ISSN: |
2444-8656 |
| DOI: |
10.2478/amns.2022.2.0126 |
| Popis: |
We apply discrete Fourier transforms to musical tempo. Then we perform pitch saliency-based melody extraction for each frame. Secondly, we use the finite element differential mathematical model for speech tracking with non-fixed parameters. In this paper, the masking threshold of each edge of the speech signal is dynamically determined by introducing a Bayesian tangent shape model. The simulation results show that the finite element differential mathematical model can remove the noise in the speech signal to a certain extent. In the environment of a low signal-to-noise ratio, the advantages of this algorithm over other spectral subtraction methods are more significant. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
