You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now the only information that goes out of users' computers is for performing speech-to-text and text-to-speech using Google services. We should do both offline.
Picovoice already released a superb model to do speech-to-text offline (https://github.com/Picovoice/cheetah), but currently, it only supports Linux. I'll try to contact them for covering more operating systems.
What about text-to-speech?
The text was updated successfully, but these errors were encountered:
Right now the only information that goes out of users' computers is for performing speech-to-text and text-to-speech using Google services. We should do both offline.
Picovoice already released a superb model to do speech-to-text offline (https://github.com/Picovoice/cheetah), but currently, it only supports Linux. I'll try to contact them for covering more operating systems.
What about text-to-speech?
The text was updated successfully, but these errors were encountered: