nlps-audio-only
Enables Dragon Voice for Krypton-only recognition (raw recognition with no semantic interpretation).
Value |
True or False DEFAULT: False (disabled) |
How to set |
Shown in precedence order (see Rules of parameter precedence): In the application’s session.xml, in the <speechrecog> section of <speechserver> under <sessionparameters>. |
Usage |
Set to True when using Krypton for raw recognition results. For example, when sending those results to a third-party semantic engine instead of using Dragon Voice for semantic intepretation. |
This parameter enables Krypton-only recognition at the session level. To set at the service level, use server.nlps.audioOnly.
Note: You cannot change between Krypton-only and semantic recognition during a single session.
For an overview, see Dragon Voice features.