Abstract
In the speech-quality evaluation of transmission systems, the selection of a sample of speakers is often a problem. This study derives relationships between speakers on the basis of their perceptual similarities and differences. A 20-item semantic differential scale was developed to yield maximum discrimination between speech from different speakers. The ratings on these scales were found to have high reliability over samples from various populations of raters and over time. Ratings were obtained on voices of 24 male and 23 female speakers. The distances between the semantic differential profiles of these speakers were calculated. These distances were then used as the input for a computational method that yields clusters of the profiles. This method, called hierarchical clustering, orders the voices in a treelike structure from a two-voice cluster to a cluster containing the total sample. The obtained branch heights in the tree are a scaled measure of the relationships between the speakers. The results of the study indicate the technique to be a promising method for the selection of a sample of speakers with perceptually different voices.
Sign-in/Register to access full text options 
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 callDisclaimer: 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.
