STT using websocket in streaming mode

Would someone please share a flow with following requirements

  • Open socket
  • Open microphone (by click)
  • Speak something
  • Watson STT transcribes
  • display interm results
  • display final result with confidence score
  • Close microphone (by click)

Thank you


Well, it is for sure close to Christmas but still, this is a place where you can get advice and help on stuff you (try to) make yourself. Like building a flow and you are almost there, just need the final help

To me, as you put your requirements, it sounds more like you should hire a consultant. Or better & cheaper, begin your own journey with Node-RED. A lot of guidelines and other documents are available for exactly that purpose

Well you could start by playing with the dashboard ui microphone node - node-red-node-ui-microphone (node) - Node-RED
It can use the built in browser STT and a couple of modes of stop start
(It can also send segments you could forward to watson but that is somewhat more complex)