Simspeech

Setting up Simspeech

The Simspeech Component allows a user to input text and have that text be interpreted by the rest of the architecture as words spoken to the robot. This offers a handy text-based alternative to spoken language.

To add the Simspeech Component to your DIARC configuration file, add:

createInstance(com.simspeech.SimSpeechRecognitionComponentImpl.class,
            "-config robot.simspeech -speaker human -listener robot");

Replace the path to the robot.simspeech file here with whatever simspeech file you wish to use. When the configuration is run, any utterances included in this simspeech file will appear in a GUI as possible buttons to click on, for convenience. However, there is also a textbox included in the GUI which the user can type anything into.

Replace "human" with the name of the intended speaker, and "listener" with the name of the intended robot, in order to set defaults for both of these fields. If you want more than one person to be speaking to the robot, then add another instance of the component for each additional person (think of these components as microphones for each human).

hello robot one
please stand
raise your arms
crouch down
stand up
please relax

Using the robot.simspeech template above, try running the config and using the buttons of the simspeech GUI to submit text utterances to the robot. If you type something new into the text box, you may notice that the robot will respond "I do not understand that." To make sure the robot understands the text input, you will need to edit the Parser dictionary.