Update: This bug has been confirmed, will be fixed on next release date. Thanks.
Regards,
Yutong
This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
We're using MS custom speech for german language in our application. When the user says the decimal number "1653,57" (in german: "Eintausendsechshundertdreiundfünfzig Komma fünf sieben") the MS speech service recognizes "1600 53,57" which are two numbers. Saying the number digit by digit doesn't help either; this results in "165 3,57" which are two numbers too.
The results are nearly the same when sayin "point" instead of "comma": "1600 53.57".
We have not tested it using a speech model for the english language.
I think this is an error in the baseline model. What can we do to be able to recognize decimal numbers?
Update: This bug has been confirmed, will be fixed on next release date. Thanks.
Regards,
Yutong
Thanks for pointing out this issue. Forwarded to product group for more investigation.
Regards,
Yutong
Thanks alot. Is there already a release date, just that we know when it will work?
@Andreas Lange Hello, the ETA product team shared to me is Sep.10th. Thanks.
Regards,
Yutong