Edit

Windows speech package — MRTK3

  • Article

Keyword Recognition

The Windows Speech package adds WindowsKeywordRecognitionSubsystem to your project, which offers keyword recognition capabilities on Windows and UWP platforms. As a MRTK KeywordRecognitionSubsystem, the subsystem can work with SpeechInteractor to trigger select events on StatefulInteractable's based on the settings of the interactables. You can also register arbitrary UnityAction's to a keyword of your choice so that the action will be invoked when such word is said.

For general information on KeywordRecognitionSubsystem in MRTK, refer to the documentation.

Setup and Usage

Please refer to the Setup and Using KeywordRecognitionSubsystem sections of the KeywordRecognitionSubsystem article. For WindowsKeywordRecognitionSubsystem, a configuration asset is used to set the Confidence Level for interpreting speech. The only capability needed in player settings is "microphone".

Dictation

The Windows Speech package also adds WindowsDictationSubsystem to your project, which offers dictation capabilities on Windows and UWP platforms. As a MRTK DictationSubsystem, the subsystem allows you to start and stop a dictation session, and provides the following events:

  • Recognizing is triggered when the recognizer is processing the input and returns a tentative result.
  • Recognized is triggered when the recognizer recognized the input and returns a final result.
  • RecognitionFinished is triggered when the recognition session is finished and returns a reason.
  • RecognitionFaulted is triggered when the recognition is faulted (i.e. error occurred) and returns a reason.

For general information on DictationSubsystem in MRTK, refer to the documentation.

Setup and Usage

Please refer to the Setup and Using DictationSubsystem sections of the DictationSubsystem article. For WindowsDictationSubsystem, a configuration asset is used to set the Confidence Level for interpreting speech and the Auto Silence Timeout and Initial Silence Timeout for a dictation session. The only capability needed in player settings is "microphone".

Text To Speech

Lastly, the Windows Speech package adds WindowsTextToSpeechSubsystem to your project, which offers the capability to synthesize and speak a text phrase on Windows and UWP platforms. For general information on TextToSpeechSubsystem in MRTK, refer to the documentation.

Setup and Usage

Please refer to the Setup and Using TextToSpeechSubsystem sections of the TextToSpeechSubsystem article. For WindowsTextToSpeechSubsystem, a configuration asset is used to select the Voice used. No capabilities are required in player settings.

Note

As the names suggest, WindowsKeywordRecognitionSubsystem, WindowsDictationSubsystem, and WindowsTextToSpeechSubsystem only work on the Windows standalone and UWP platforms.