macaroni

recognize

Description:
The recognize executor is used to recognize audio from a device or a file, using the Azure Speech SDK.

Parent: execute, then, else

Value (simple form): string specifying a phrase or list of phrases to improve recognition accuracy (default: null)
Value (key/value form):

• key: string, optional, specifying the Azure Speech api key (default: from configuration)
• region: string, optional, specifying the Azure Speech region (default: from configuration)
• phrases: string or sequence of strings, optional, specifying the phrase or list of phrases to recognize (default: null)
• file: string, optional, specifying the input file name (default: null)
• once: boolean, optional, specifying if only one utterance should be recognized, or more than one
  • for device, default is true
  • for file, default is false
• timeout: time, optional, specifying the amount of time, after which the recognize will be stopped
  • if timeout has no unit, milliseconds is assumed
  • for device, default is 30s
  • for file, default is null
• After the user speaks a phrase, subsequent executors in the same execution context can access the recognized result:
  • recognize.text - the text recognized
  • recognize.result - the Speech SDK SpeechRecognitionResult object

# simple form
recognize: (string)

# simple example
recognize:

# key/value form
recognize:
  key: (string, optional)
  region: (string, optional)
  phrases: (string or sequence of string, optional)
  once: (boolean, optional)
  file: (string, optional)

# key/value example
recognize:
  file: `{folders.desktop}\audio.wav`