Azure AI Speech
An Azure service that integrates speech processing into apps and services.
1,519 questions
Battalion is misspelled in the speech-to-text return
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(51.2, 78%, 14%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EOA%3C/text%3E%3C/svg%3E)
Owen Allen
1
Reputation point
I'm working on audio data files that have a lot of radio chatter between fire departments. The word "battalion" often comes up, and the speech-to-text misspells it every time as "battallion".
Here is an example audio file with multiple utterances https://storage.googleapis.com/fire-brothers-public/1608251646578-1.wav . I'm unable to find a place to submit a proper bug report for this. The twitter handle @AzureSupport told me to post here.
{count} votes
-
romungi-MSFT 43,656 Reputation points Microsoft Employee2020-12-18T12:06:41.827+00:00 @Owen Allen I have tried to replicate this scenario with my own voice using the JavaScript SDK hosted on demo page. The result from the service seems to be returning the correct spelling when speaking from the microphone.
When I try the same file the text seems to recognize the word battalion as Italian, so I think the radio chatter in this audio might be causing the issue. If you are using any of the SDK and noticed this issue the right place to report this is the speech SDK github repo. We will also pass this feedback to our team internally to get feedback on the audio and the response. Thanks for reporting.
Sign in to comment