Speech and Language Technology for Disabled Persons

Stockholm, Sweden
May 31 - June 2, 1993

Personalizing Synthesized Voices

Debra Yarrington, Richard Foulds

Applied Science and Engineering Laboratories, University of Delaware/ A.I. duPont Institute, Wilmington, DE, USA

At the Applied Science and Engineering Laboratories (ASEL), a method is being developed for creating personalized synthesized voices that are unique and appropriate for each individual with severe speech impairments. A frequent complaint heard among users of augmentative communication devices is that the number of voices available is extremely limited. Often, the user is forced to communicate with a voice that is completely inappropriate for that individual. The ASEL method creates synthesized voices that are not only age and gender appropriate, but also have the appropriate accents and dialects included in the voices. At the same time, these voices are very intelligible and allow for an unlimited vocabulary.

Full Paper

Bibliographic reference.  Yarrington, Debra / Foulds, Richard (1993): "Personalizing synthesized voices", In SLTDP-93, 169-172.