Vocabulary Synthesis Based on Line Spectrum Pairs
Abstract
This report describes the initial investigation of a new synthetic speech system based on a line spectrum pair (LSP) representation of the speech spectral envelope. The system contains a library of stored LSP speech segments extracted from natural speech. These segments are modified as necessary by a small set of context-sensitive rules and then concatenated to generate high- quality, natural-sounding speech. Tests of a preliminary system produced Modified Rhyme Test and Diagnostic Rhyme Test scores of 87.3 and 79.7, respectively. This LSP vocabulary synthesizer is currently limited to single syllable utterances, but future research will expand its capabilities to allow implementation of a full text-to-speech system.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 12, 1989
- Accession Number
- ADA203998
Entities
People
- Stephanie S. Everett
Organizations
- United States Naval Research Laboratory