0

i need to implement speech recognition client side. I implemented with html5 webkitSpeechRecognition but it only works on chrome browser. Is there any other API which will work on all broswers?

user1457957
  • 195
  • 1
  • 6
  • 21
  • Possible duplicate of [Cross Browser Speech Recognition](http://stackoverflow.com/questions/25789249/cross-browser-speech-recognition) – Nikolay Shmyrev Apr 05 '17 at 10:42

1 Answers1

1

you can try the Watson STT APIs, they work not only in Chrome, but in Firefox and many versions of Safari. Try the demo here: https://speech-to-text-demo.mybluemix.net/

Daniel Bolanos
  • 770
  • 3
  • 6
  • Watson STT java sdk will do the speech recognition at server side. I do not have speakers on server, So need to capture the voice from client side and pass the converted the text to watson conversation API. – user1457957 Apr 05 '17 at 18:16
  • you can feed audio from the browser to Watson STT and then feed the stream of hypotheses that come to you in real time to the Watson conversation API – Daniel Bolanos Apr 05 '17 at 21:08
  • yes i did that. Captured the audio and send .wav file to server from there to STT and from there to conversation API, but it is very slow due to so much processing in middle. – user1457957 Apr 05 '17 at 21:38
  • 1
    please do not do it file based, but streaming the audio, otherwise latency will be an issue for sure. Please take a look at this: https://github.com/watson-developer-cloud/speech-javascript-sdk – Daniel Bolanos Apr 06 '17 at 22:05
  • those examples are in node js. I am using jquery and angularjs – user1457957 Apr 07 '17 at 19:30