How do I extend the waiting time of Azure speech-to-text API in Python?

Question

When using speech-to-text to transfer audio file to text, I found that the function would stop working if human voices haven't occurred for about 5 seconds. In my case, what I want to transfer is audios of interviews, which would often contain some advertisements or music in the middle of it, and when this happens, the speech-to-text would only transfer the first half of the whole audio, and report an error that "No speech could be recognized".
In this case, how can I extend the waiting time of that in order to transfer the whole file in Python codes?

Answer 1

Hello @Muyao Hu

I think there are two solutions you can have a try in Python SDK:

1. There is a 'set_property' method on the config to allow you to set parameters to your request, which can change the default silence time:: https://learn.microsoft.com/en-us/python/api/azure-cognitiveservices-speech/azure.cognitiveservices.speech.propertycollection?view=azure-python#azure-cognitiveservices-speech-propertycollection-set-property
2. This way you can set the EndSilenceTimeout (PropertyIDs in Pyhton: https://learn.microsoft.com/en-us/python/api/azure-cognitiveservices-speech/azure.cognitiveservices.speech.propertyid?view=azure-python#fields)

Please notice, the time is as "ms". Hope above helps!

Regards,
Yutong

-Please kindly accept the answer if you feel helpful, thanks!