search

microsoft / Cognitive-SpeakerRecognition-Windows

Windows SDK for the Microsoft Speaker Recognition API, part of Cognitive Services
https://www.microsoft.com/cognitive-services/en-us/speaker-recognition-api
Other
60 stars 62 forks source link

SpeechNotRecognized when recorded in browser #11

Open youngsun0 opened 5 years ago

youngsun0 commented 5 years ago

I recorded my voice in my Chrome browser in the correct format (16 bit, 16khz mono) and I get a speechNotRecognized error. However, when I export the Wav file in Audacity without changing anything, my voice enrolls perfectly. The two recordings sound identical so I am confused as to why one gets accepted and not the other.

DevSum commented 5 years ago

I meet that error in the brower, too

youngsun0 commented 5 years ago

I would suggest looking at the RIFF header of the wav file that youre sending

DevSum commented 5 years ago

Does it works well by RESTful API? I'm in a project and I just tried the brower.

youngsun0 commented 5 years ago

Yep. What format is the audio in when you send it?

DevSum commented 5 years ago

I didn't send files..., I just tried record on the brower.

youngsun0 commented 5 years ago

You must have sent something over to Microsoft for you to get that error

DevSum commented 5 years ago

I tried this page: https://azure.microsoft.com/services/cognitive-services/speaker-recognition/