Indian English model does not perform well in far-field scenario [202760222]

Links (3)

“

covering multiple language recognition, that is, in the config of the transcription request besides the main language code that STT should aim to transcript to (the one specified in the languageCode field) you can also provide alternative language codes in case the transcription of the main one fails (specified in the alternativeLanguageCodes fields, as demonstrated in the documentation provided). In your case, you should specify en-IN as the main language code, and en-US as an alternative one. I believe that could be a good solution for your use case, but since I’m not sure about the nature of your audio files, I can’t jump into conclusions. ”

be...@

Issue 202760222

Links (3)

Issue metadata