UCLA and MSR Announce
Release of Formant Trajectory Corpus
By
LI DENG & ABEER ALWAN
The speech group at Microsoft
Research, Redmond and IPAM at UCLA have recently jointly developed a database
for manually labeled vocal-tract-resonance or formant trajectories, for research
in speech processing including analysis, synthesis, and recognition. The
database contains a representative subset of the TIMIT corpus with respect to
speaker, gender, dialect and phonetic context, with a total of 538 sentences.
The data can be downloaded
by going to
http://www.icsl.ucla.edu/~spapl and then going to the
Free Software and Shareware section.
The download consists of
data and user manual, as well as an ICASSP-2006 paper describing technical
details of the database.
