1. Linguist List
  2. espeak_yid

Overview

HTTPS SSH

Yiddish Espeak Module

This is a first version of an Espeak module for Yiddish. This is pre-alpha development that we use to create a model that can be used as a module in Praat for automatic or forced alignment of transcription of Yiddish recorded speech. Praat offers the functionality to align transcribed text to the speech signal along the time axis. It uses analysis via synthesis as a method. The transcribed text is used to generate an audio signal that is then matched against the audio signal in the recording to generate hypotheses about time boundaries. (see TODO cite paper!)

The initial module was developed by Andrew Lamont (UMass, Amherst) and Lwin Moe (Indiana University). It is part of the Yiddish speech projects at Indiana University.

This sub-project is part of the larger GORILLA project managed by Damir Cavar and Malgosia Cavar. It is also related to Dov-Ber Kerler's AHEYM project.

Instructions

This section is under development. We are working on more detailed instructions.

Compiling Espeak

./configure
make
ESPEAK_DATA_PATH=`pwd` LD_LIBRARY_PATH=src:${LD_LIBRARY_PATH} src/espeak "How are you?"

Adding a new language

./mkdictlist Makefile.am
make my

ESPEAK_DATA_PATH=`pwd` LD_LIBRARY_PATH=src:${LD_LIBRARY_PATH} src/espeak --voices=my
ESPEAK_DATA_PATH=`pwd` LD_LIBRARY_PATH=src:${LD_LIBRARY_PATH} src/espeak -v my "စပေါ်"