Hello,
Today, we tried to analyze Video speech using the "Pronunciation Assessment". But, at one time, we were allowed to analyze only 20 seconds' videos.
How we can analyze the entire video (about 3 minutes)?
Hello,
Today, we tried to analyze Video speech using the "Pronunciation Assessment". But, at one time, we were allowed to analyze only 20 seconds' videos.
How we can analyze the entire video (about 3 minutes)?
Hello,
Is there any code sample you are referring to for your project? We have a video to explain the sample code here https://www.youtube.com/watch?v=zFlwm7N4Awc .
Hope this helps.
Regards,
Yutong
Hello,
I have tried again on my side and it works for me well. As below screenshot, it's successfully with my 30sec audio. This is the sample code repo I am using, all of them are very convenient. https://github.com/Azure-Samples/Cognitive-Speech-TTS/tree/master/PronunciationAssessment/CSharp/Console

Please let me any block for this and share the code sample you are using.
Regards,
Yutong
Hi, @JunbumKwon-2990
To handle long speech for pronunciation assessment, you can refer to below sample code:
https://github.com/Azure-Samples/cognitive-services-speech-sdk/blob/master/samples/python/console/speech_sample.py#L643
It is based on continuous recognition and therefore it doesn't have limitation on length.
Regards,
Yinhe
6 people are following this question.