Open OSS542 opened 3 years ago
Yes! I have this partially working already. For open transcription, it works just fine. But for pre-trained voice command recognition, I've only found the ability to restrict the vocabulary. I may have to dig deeper to get grammar-based recognition working.
Were you planning to use Vosk for open transcription or restricted voice commands?
At present I plan on using Vosk for open transcription, as it is better suited to that at present. I use PocketSphinx for restricted vocabulary as it is considerably faster at this, and the vocabulary is easy to modify.
Is there any chance of adding support for Vosk to voice2json ?