Abstract
AbstractThis paper considers the use of the ARMA model in speech synthesis and discusses the quality improvement and problems in synthesized speech. First, the construction of the analysis filter is considered which is used to determine the residual signal in the speech synthesis by the ARMA model. It is pointed out that special considerations must be given to the MA part. Then the qualities of the speech synthesized by ARMA coefficients and the speech synthesized by AR coefficients are compared. the quality of speech was evaluated based on the preference measure in the listening test (Sheffé's paired comparison). It is shown that the quality of the synthesized speech is better when ARMA coefficients are employed. It is also shown that the quality of the synthesized speech in the AR model is not affected greatly by the method of analysis. Finally, a new measure for the spectral distortion is proposed as a measure for objective evaluation of the quality of the synthesized speech. the measure considers the properties of the MA part of the ARMA model and gives a result which corresponds well with the result of evaluation by the listening test.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Electronics and Communications in Japan (Part III: Fundamental Electronic Science)
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
