Speech-to-text from a LIVE stream

Question

I have a Java application and I want to transcribe, not a file, but a LIVE stream(rtmp:// or m3u8) that is being served by Wowza.

Is that possible?

Thank you

PS: I'm willing to pay any of those cloud services that offer audio transcription, but the ones I've seen don't seem to have this functionality.

Hi RichardW, Watson STT does not support this out of the box, but this could be accomplished on your end by implementing a bridge rtmp <> websockets. What volume of speech do you plan to push to the STT service? — Daniel Bolanos, Jul 20 '16 at 16:57

score 5 · Answer 1 · edited May 23 '17 at 11:51

There are a few options!

Google has one that is in beta, but seems very promising! Their documentation is a little light at this point but here is a good github example. https://github.com/GoogleCloudPlatform/java-docs-samples/tree/master/speech

and their documention on it:

Best of luck! If you end up trying out both I would love to have an update on which one you felt worked best!

EDIT Another useful link: Stream audio from mic to IBM Watson SpeechToText Web service using Java SDK while it doesn't interface with Wowza, this might be a good place to start for IBM.

1 Answers1