Azure AI Speech Demo: Continuous Transcription Insights (Full Transcript)

Speaker 1: In this demo, you'll see how to use Azure AI Speech to perform continuous transcription using the JavaScript SDK. For further details on how this code hangs together, you can visit the blog at jamiemcguire.net. So what we're looking at here is a simple HTML page, and we can see we've got a bunch of fields. We've got the subscription ID, which is taken from the Azure portal, and we've also got the region that this particular Azure AI Speech service is provisioned in. We've got a couple of buttons, one to start the recognition and another to stop the recognition. When we click start recognition and start to speak into the microphone, what we'll see is down at the console at the bottom of the browser, we'll start to see text being identified from the speech and converted, or rather displayed. We'll also see that same text start to be rendered in the results text box. So we'll do that just now. We'll click start recognition. We get prompted to provide access to the microphone. So we'll click allow. After clicking allow, we can see down towards the bottom here that the audio is starting to be transcripted directly from speech into text. We can also see the results in the text box starting to fill up. We can scroll down here and we can start to see more text being identified as I speak. So there's quite a few different use cases for this. The main use case for this at the minute for me personally is for the AudioNotes web application that I'm going to be building in the next week or so. We can click stop recognition. You can see that the session stopped and the text is no longer transcribed from the audio. So that was a very, very quick run through of Azure AI Speech and how to use the JavaScript Client SDK to perform continuous transcription. You can learn more about how this code has been written and hangs together by looking at the accompanying blog, which is detailed in the description for this video. Thank you.