Recognition Mode: Automatic
<listen mode="automatic">
Remarks
Automatic mode is used for recognitions in telephony or hands-free scenarios. For an automatic recognition, the general principle is that the Speech Platform (rather than the application) controls when to stop the recognition process. No explicit Stop method call is necessary from the application, because the end of an utterance is automatically determined using the endsilence value.
As soon as a recognition is available (the endsilence time period is used to determine the phrase-end silence that implies recognition is complete), the Speech Platform automatically stops recognition and returns its results. The onreco event is raised when a successful recognition has taken place (confidence equal to or greater than the threshold specified in the reject attribute), and an onnoreco event occurs when there is no recognition (confidence less than the threshold specified in the reject attribute).
- Case 1 illustrates a normal recognition.
- Case 2 demonstrates initiating an onnoreco event after the babbletimeout interval is exceeded, which ends a listen.
- Case 3 displays an unsuccessful recognition attempt where the recognizer initiates an onnoreco event before the endsilence interval at the end of an utterance has been exceeded.
- Case 4 demonstrates no input from the user. This results cause an onsilence event to occur.
As noted above, all events except onspeechdetected end listen in automatic mode.
The following event diagram illustrates the automatic-mode listen recognition timelines.
See Also
listen Element | Recognition Mode: Automatic Example | Recognition Mode: Single | Recognition Mode: Multiple