This folder contains Node.js code that records WAV file from the microphone, sends it to Google Cloud Speech-To-Text API for recognition, then send the resulting text to Google Cloud Text-To-Speech API for converting back to WAV file, and plays the resulting file (basically, repeating what you said, but using a Google Cloud Text-To-Speech voice).
To use this sample, you'll need to have the following:
- Raspberry Pi with Raspbian.
- Microphone and speaker.
- Google Cloud API JSON key that has access to Speech service. Point environment
variable
GOOGLE_APPLICATION_CREDENTIALSto the key file.
Actually, there is nothing specific to Raspberry Pi in this particular sample, as the code should work on any Linux with ALSA.
Before running the code, make sure to install all the dependencies by running
npm installEnjoy!