SRT from speech-to-text in Azure results in combined captiona

Saariko 20 Reputation points
2024-03-12T11:50:19.07+00:00

hey all, when trying out on Azure portal to generate a captions file (srt) i get expected results, however when doing the same from my node app using the microsoft-cognitiveservices-speech-sdk library i get the results combined.

anyone know why?

wrong results:
1

00:00:00,790 --> 00:00:17,030

Dad, you OK? It's our phone bill. We pay for things that we don't need. Loaded bundles to reckless spending, no more. That's a bit dramatic. A better plan starts at $25 a line. Did you say $25 a line? And see big on things we love streaming.

2

00:00:18,680 --> 00:00:29,560

That's awesome. Spaghetti night. Dinner in 30? Oh, happy day. A better plan to save is provider. It starts at $25 per line, guaranteed for three years and get both streamers for just $10 per month only on provider.

expected results:
1

00:00:00,790 --> 00:00:03,150

Dad, you OK?

It's our phone bill.

2

00:00:03,670 --> 00:00:08,870

You pay for things that we don't need.

Loaded bundles to reckless spending, no more.

3

00:00:08,870 --> 00:00:13,070

That's a bit dramatic.

A better plan starts at $25 a line.

4

00:00:13,070 --> 00:00:16,710

Did you say $25 a line?

And say big on things we love streaming.

5

00:00:16,710 --> 00:00:20,950

Oh, that's awesome.

Spaghetti night.

6

00:00:20,950 --> 00:00:23,070

Dinner in 30.

Oh, happy day.

7

00:00:23,070 --> 00:00:26,510

A better plan to save is provider.

It starts at $25 per line, guaranteed for three

8

00:00:26,510 --> 00:00:30,830

years and get both streamers for just $10 per month.

Only on provider.

Azure AI Speech
Azure AI Speech
An Azure service that integrates speech processing into apps and services.
1,369 questions
{count} votes

Accepted answer
  1. VasaviLankipalle-MSFT 13,611 Reputation points
    2024-03-12T22:47:23.1666667+00:00

    Hello @Saariko , Thank you again for your time and patience throughout this issue.

    We apologize for the inconveniences that caused. Looks like the product team is working on this. But there is no ETA when the Java and JavaScript samples get updated.

    You can track this issue here: https://github.com/Azure-Samples/cognitive-services-speech-sdk/issues/2071#issuecomment-1986466853

    Regards,
    Vasavi

    Please remember to "Accept Answer" if any answer/reply helped, so that others in the community facing similar issues can easily find the solution.

    0 comments No comments

0 additional answers

Sort by: Most helpful