To convert speech to text successfully, the IDOL Speech Server must know the language of the speech.
The IDOL Speech Server can automatically identify the language of speech. If language identification is not bypassed using one of the following methods it is performed automatically.
To bypass language identification
IdolSpeechLanguage
configuration parameter. This parameter specifies the language pack to use for all documents and takes precedence over other language settings. You can set this parameter when all of your audio is in the same language.To bypass language identification for a single document, add the field AUTN_AUDIO_LANGUAGE
to the document. The value of the field must identify the language pack to use for transcription, for example:
#DREFIELD AUTN_AUDIO_LANGUAGE="ENUK"
For a list of IDOL Speech Server language packs, refer to the IDOL Speech Server Administration Guide.
|