Continuous Speech Recognition with Interrupt/Barge-In Support

[asadahmedkhan]

I have a question

Hi,

I’m working on integrating continuous speech recognition in my web application using Azure Cognitive Services / Bot Framework Web Chat. My goal is to allow the user to speak continuously while the bot is responding with text-to-speech, and ideally, the user should be able to interrupt the bot’s speech by speaking (barge-in).

Here’s what I have so far:

Speech-to-text works on mic click.

Text-to-speech is functional.

Currently, when the bot is speaking, the mic is disabled, and the user cannot interrupt.

I want to achieve a flow like:

User clicks the mic → starts speaking → speech-to-text captures input.

Bot responds with text-to-speech.

User should be able to start speaking at any time, even if the bot is talking, and have the bot stop its speech and process the user input.

Questions:

How can I implement continuous speech recognition while the bot is speaking?

Is there a way to configure automatic barge-in in Bot Framework Web Chat with Azure Cognitive Services?

Can the mic and speaker be active simultaneously in the browser for this use case?

Any sample implementation or guidance would be greatly appreciated.

Thanks!