See Also: SVAPI, IdentificationNotify
public abstract IdentificationResults[] identify(Model Models[], Utterance Utterance, int MaxCount)
Performs speaker identification. The models are compared against the utterance. A subset of the models are returned with their raw scores in the IdentificationResults array. This function returns an N-best results list. The list is ordered according to raw score, with the best score first.
Parameters:
Exceptions:
public abstract void enableAsynchronousIdentification(Model Models[], int MaxCount, IdentificationNotify Notify)
Enables asynchronous identification. If enabled, the engine passively identifies the speaker. If the raw scores change significantly, or the speaker appears to have changed, the engine calls IdentificationNotify.identificationScoreChanged(). For a text-dependent engine, this would probably be after the speaker has spoken the password. An engine may also choose to call this function whenever it detects that something has been spoken.
Parameters:
Exceptions:
See Also: identificationScoreChanged, disableAsynchronousIdentification
public abstract void disableAsynchronousIdentification()
Disables asynchronous identification.
Exceptions:
See Also: enableAsynchronousIdentification
public abstract Object getPrompt(Class Type[])
Returns a prompt for the application to present to the user. The application passes to the engine an array of Class objects that represent the possible prompt object classes that it can handle. For example, suppose an application could prompt with a gui or with audio, but prefered gui. The application would create an array of two and put the Class object for Strings in the first element and the Class object for Utterances in the second.
Exceptions:
See Also: getPrompt