A method for providing selectivity to a speech-to-text system operates a voice transcription assistant coupled to the speech-to-text system such that speech detected by the speech-to-text system is not transcribed into text for display to a user unless the voice transcription assistant determines that the detected speech is relevant to the user.
Smart glasses, including at least one microphone, a speech-to-text processing unit, a display, and a voice transcription assistant, provide selective speech-to-text functionality. The voice transcription assistant has one or more processors; and logic encoded in one or more non-transitory media for execution by the one or more processors and when executed operable to prevent the speech-to-text processing unit providing text to the display in response to the reception of speech by the microphone unless the logic determines that the received speech is relevant to the user.