linear prediction

Allam's picture

Speech segmentation in synthesized speech morphing using pitch shifting

Journal Title, Volume, Page: 
The International Arab Journal of Information Technology, Vol. 8, No. 2, April 2011
Year of Publication: 
2011
Authors: 
Allam Mousa
Department of Telecommunication Engineering, Faculty of Engineering and Information Technology, An-Najah National University, Nablus. Palestine
Current Affiliation: 
Department of Telecommunication Engineering, Faculty of Engineering and Information Technology, An-Najah National University, Nablus. Palestine
Preferred Abstract (Original): 

This paper discusses the speech morphing process showing some limitations of using the directly obtained LPC and excitation parameters of speech. The algorithm here depends on changing the pitch of the source to match that of the target based on analyzing the speech signals to its basic components. Different experiments for changing the female to female, male to male, male to female and female to male speech were performed. Interesting results were obtained while dealing with children's speech. Difficulties of obtaining the pitch period were overcome but the obtained results have some diversity in the quality of performance even though the pitch has been changed correctly. The method for obtaining LPC and excitation used could be improved which could provide better results for this application.

Syndicate content