Transcribing audios just from my microphone #117
Replies: 1 comment
-
Hi @filipe-campos The Web Speech API will only transcribe audio from your microphone (I'm assuming your microphone was picking up other voices from your computer speakers). The ability to distinguish between speakers depends on the implementation of the speech-to-text API on the backend (on Chrome, this is Google's speech-to-text cloud service) - The Web Speech API specification does not include any fields for multiple voices - even if a cloud service did support it, we would need to add something custom to this library to support something off-spec like that. So I don't think there's a good solution for this yet, at least not for web apps. If this is a critical requirement for your application, you may need to do the following:
It's quite a lot of work! I have a few things on my TODO list that could support the above one day:
|
Beta Was this translation helpful? Give feedback.
-
The library is transcribing the audios from my microphone and also from other people if I'm on a call. I would like to just transcribe the audio from my microphone.
Does anyone have any idea if this is possible?
Beta Was this translation helpful? Give feedback.
All reactions