Microsoft Speech Platform

---

ISpRecognizer2::SetTrainingState

ISpRecognizer2::SetTrainingState puts the recognizer into training mode to do acoustic adaptation. This is what is happening in the background when you are using the training wizard from the control panel.

When in training mode, the recognizer assumes that what the user said is definitely correct and that it should definitely be used for adaptation. The assumption is that the user is reading from a script and there are no forks in the script.

HRESULT SetTrainingState(
   BOOL    fDoingTraining,
   BOOL    fAdaptFromTrainingData
);

Parameters

fDoingTraining

True to switch training on, false to switch it off.

fAdaptFromTrainingData

Tells the engine whether to actually use the data or discard it. This is only really useful at the end of training (that is, when fDoingTraining is set to false).

Return values

Value	Description
S_OK	Function completed successfully.