Hello,
We are trying to use voice-to-text service to transcribe incoming call from a phone. The call is picked up by a virtual softphone and the analogue signal is digitized using PCM. Then the data packet is send to Watson voice-to-text to be converted with a narrow band mode used. However, often times we don't get any script output. Are we missing something here?
Thanks for you help in advance.
Flora
Do you have a control in your sample test usage. IE. Are you capturing the audio to verify what has been captured?
Hello @FQian can you provide more details on what is going on? Are you using the REST or the WebSocket interface?
please make sure you use the flag continuous=true for continuous recognition. Otherwise the recognition will stop after the first utterance.
Dani
Answer by FQian (1) | Oct 27, 2016 at 12:33 PM
@chughts and @Daniel Bolanos, thanks for your reply. We use WebSocket. Here's the specific piece of code we used. After turned the flag continuous=true, we are still not getting output text.
Answer by Daniel Bolanos (1252) | Oct 28, 2016 at 11:03 AM
Hi @FQian are you using the Java SDK?
Have you tried with the examples provided?
If that does not work for you, please open an issue in that repo.
Dani