A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.
Zaki, F. (2021). A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.. EKB Journal Management System, 15(1), 1-13. doi: 10.21608/bfemu.2021.170547
Fayez W. Zaki. "A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.". EKB Journal Management System, 15, 1, 2021, 1-13. doi: 10.21608/bfemu.2021.170547
Zaki, F. (2021). 'A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.', EKB Journal Management System, 15(1), pp. 1-13. doi: 10.21608/bfemu.2021.170547
Zaki, F. A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.. EKB Journal Management System, 2021; 15(1): 1-13. doi: 10.21608/bfemu.2021.170547

A New Adaptive Line Spectral Pair Filter for Linear Prediction of Speech.

MEJ- Mansoura Engineering Journal
Article 5, Volume 15, Issue 1, June 1990, Page 1-13  XML PDF (123.38 K)
Document Type: Research Studies
DOI: 10.21608/bfemu.2021.170547
View on SCiNiTO View on SCiNiTO
Author
Fayez W. Zaki email
Department of Engineering Mu'tah University, Jordan.
Abstract
 Analysis/synthesis of speech by PARCOR method is proved to be advantageous over other methods previously developed for bit rates above 2.4 kbps, However, if the bit rate is reduced below this value, reconstructed speech becomes unclear and unnatural. Line spectral pair (LSP) coefficients were thus Introduced as alternative representations of linear prediction parameters for low-bit rate speech coding. This paper presents a least-mean-square (LMS) type adaptive LSP filter for directly calculating the LSP coefficients on sample-by-semple basis. The algorithm proposed here is shown to have higher convergence rate and lower misadjustment as compared to other algorithms. 
Statistics
Article View: 56
PDF Download: 109