ASRWavesurferPlugin

Automatic Speech Recognition Plugin for Wavesurfer

Reference

The WaveSurfer Automatic Speech Recognition Plugin
Giampiero Salvi and Niklas Vanhainen
Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC'14), May 2014, Reykjavik, Iceland.

Additional Resources

You can find the acoustic and language models required for the plugin to work at this address:

http://www.speech.kth.se/asr/

Installation

Preliminaries (all platforms):

• Install WaveSurfer from http://wavesurfer.sourceforge.net/
• Run WaveSurfer once and then close the program (this creates the preference directory)

Installation Linux/Mac OS X:

• unpack v1.0.2.tar.gz into ~/.wavesurfer/1.8/plugins/ with:

tar xvfz v1.0.2.tar.gz -C ~/.wavesurfer/1.8/plugins/

• unpack at least one of the language packs (from http://www.speech.kth.se/asr/) into ~/.wavesurfer/1.8/plugins/asr/ with:

tar xvfz wavesurfer_asr_plugin_Swedish_models_0.1.tgz -C ~/.wavesurfer/1.8/plugins/asr/
tar xvfz wavesurfer_asr_plugin_English_models_0.1.tgz -C ~/.wavesurfer/1.8/plugins/asr/

Installation MS Windows:

• unpack v1.0.2.tar.gz into ${HOME}\.wavesurfer\1.8\plugins\
• unpack at least one of the language packs into ${HOME}\.wavesurfer\1.8\plugins\asr\ Where ${HOME} is usually under C:\Users\<username>