Loading...
Please wait, while we are loading the content...
Similar Documents
Analysis and improvement of the vector quantization in selp (stochastically excited linear prediction)
| Content Provider | NASA Technical Reports Server (NTRS) |
|---|---|
| Author | Ketchum, R. H. Kleijn, W. B. Krasinski, D. J. |
| Copyright Year | 1988 |
| Description | The Stochastically Excited Linear Prediction (SELP) algorithm is described as a speech coding method employing a two-stage vector quantization. The first stage uses an adaptive codebook which efficiently encodes the periodicity of voiced speech, and the second stage uses a stochastic codebook to encode the remainder of the excitation signal. The adaptive codebook performs well when the pitch period of the speech signal is larger than the frame size. An extension is introduced, which increases its performance for the case that the frame size is longer than the pitch period. The performance of the stochastic stage, which improves with frame length, is shown to be best in those sections of the speech signal where a high level of short-term correlations is present. It can be concluded that the SELP algorithm performs best during voiced speech where the pitch period is longer than the frame length. |
| File Size | 367793 |
| Page Count | 6 |
| File Format | |
| Alternate Webpage(s) | http://archive.org/details/NASA_NTRS_Archive_19880016372 |
| Archival Resource Key | ark:/13960/t1wd8rk8w |
| Language | English |
| Publisher Date | 1988-05-01 |
| Access Restriction | Open |
| Subject Keyword | Communications And Radar Coders Linear Prediction Algorithms Coding Vectors Mathematics Speech Counters Voice Communication Ntrs Nasa Technical Reports Server (ntrs) Nasa Technical Reports Server Aerodynamics Aircraft Aerospace Engineering Aerospace Aeronautic Space Science |
| Content Type | Text |
| Resource Type | Article |
