TR-IT-0206 :1997.03

Harald Singer, Masahiro Tonomura, Qiang Huo, Jun Ishii, Toshiaki Fukada, Michael Schuster

Baseline Acoustic Models for the Spoken Language Database (SDB/SLDB)

Abstract:In this document we describe the creation of speaker-independent and gender-dependent acoustic models from ATR's speech and language (SLDB) and speech databases (SDB) both for phoneme context-independent and context-dependent case. Models and parameter files have been installed in an accessible file system inside ATR ITL. The software used was ATRSPREC version r04a10. This report can be accessed online via

・http://www.itl.atr.co.jp/~singer/publication/TR-IT-0206/baseline.ps.gz

・http://www.itl.atr.eo.jp/~singer/publication/TR-IT-0206/baseline/baseline.html