Example for connecting real-time streaming through socket.io to a Node server

An easy-to-set-up playground for cross device real-time Speech Recognition amd intent extraction with a Node server and socket.io.

NLU Options

1. Using Google Cloud Speech for transcription and DialogFlow for extracting intent from the trnscription
2. Using DialogFlow audio streaming intent detect to transcribe and extract intent in one call

How to run

1. Set a service account for Google Cloud Speech
2. Set a service acconut for Dialogflow project
3. For each service acconut, set their private_key and email fields values in .env
4. Choose the Google Cloud Speech option by setting 'NLU' in .env to 'GOOGLE_SPEECH. For Dialogflow option, don't set the value

Made by Zach (Tzahi) Efrati

This is a work in progress. Please feel free to submit issues and PR!

Work is based on the GitHub repository by Vinzenz Aubry