How to proceed multiple streamingRecognize request in Google Speech-to-Text API?

[Tom Chiang]

I have built a web based Speech-to-Text application with Google Speech-to-Text API.

The application flow:

1. Get audio stream from browser 
   
2. Stream audio data to a speech recognition service and get real time results as our user talks

I have referenced the sample code in my application, the streamingRecognize request works perfectly with a signal client call, and my server can receive the interim-transcription result from Google Speech API. However, when there are more then one client call streamingRecognize simultaneously, Google Speech API will get stuck. So, may I know how to handle more than one streamingRecognize request concurrently? Do I need to create another client to process the second streamingRecognize request? Thanks.

   

Back-end server specifications:

• NodeJS
• Socket.io

[Fabio Santos]

Wow, this is pretty recent, I'm also on the same boat right. Everything working flawlessly except for this, I might have to create a JS Object like {[userKey]: new speech.SpeechClient} as to add all users as they need to talk.

[Tom Chiang]

Umm... So it this the only way to archive the goal by implementing like this?

eval("var " + myString + " = new speech.SpeechClient();");

Fabio Santos於 2020年2月21日星期五 UTC+8上午1時58分36秒寫道：