ENVOICE is a concatenative speech synthesis system developed by
the SLS group. ENVOICE creates synthetic speech by concatentating
segments of speech from a pre-recorded speech corpus. The
concatentation of segments can occur at the phrase, word, or sub-word
levels. The choice of segments to concatenate is determined by
searching the pre-recorded corpus to find the segments best matching a
set of concatenation constraints. ENVOICE takes advantage of
finite-state transducer (FST) technology to efficiently deliver
synthetic speech on-demand.
Natural-Sounding Speech Synthesis Using Variable-Length Units
M.Eng. thesis, MIT Department of Electrical Engineering and Computer Science, May 1998.
J. Yi and J. Glass,
"Natural-Sounding Speech Synthesis Using Variable-Length Units,"
Proc. ICSLP, Sydney, Australia, November 1998.
J. Yi, J. Glass and L. Hetherington,
"A Flexible, Scalable Finite-State Transducer Architecture
for Corpus-Based Concatenative Speech Synthesis,"
Proc. ICSLP, Beijing, China, October 2000.