kadosu-dev team mailing list archive

[Martin Schaaf <mascha@xxxxxxxxxx>]

Blueprint changed by Martin Schaaf:

Whiteboard changed to:

Sphinx4 is used as speech recognition lib, because it is a pure java
implementation. Voxforge site has acoustic models for Sphinx. Input is a
wav file. All audio files need to be converted to wav. There is a sample
implementation for recognizing speech from wav files in Sphinx that
takes about 20 lines of source code.

-- 
  Indexing audio
  https://blueprints.edge.launchpad.net/kadosu/+spec/index-audio-files