Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Support external speech to text services #197

Open
metronidazole opened this issue Oct 5, 2023 · 12 comments
Open

Support external speech to text services #197

metronidazole opened this issue Oct 5, 2023 · 12 comments
Labels
enhancement New feature or request tts&stt Speech-to-text, text-to-speech and wakeword requests or bugs, including Vosk

Comments

@metronidazole
Copy link

It would be nice if this supported android's platform for speech to text services so I could use an external app for voice to text. e.g. FUTO voice input https://www.youtube.com/watch?v=UCGaKvZpJYc

@metronidazole metronidazole changed the title Support external speech to services Support external speech to text services Oct 5, 2023
@Stypox
Copy link
Owner

Stypox commented Oct 6, 2023

Thank you for pointing that out to me! That's a fantastic advancement for speech to text on Android!

@paolo-caroni
Copy link

paolo-caroni commented Oct 25, 2023

I have added the repo to my fdroid, thanks @metronidazole

Also there is more issue like this... it's not the first time that is asked and it is on the roadmap.

@paolo-caroni
Copy link

I would like to add that seems that FUTO is not yet a really FOSS project, also in not listed as a vocal input, it support only IME API.

@AyoungDukie
Copy link

For reference, this still appears to be the case that Futo is using their FTL "source-available" license. However, at least in Android 14, it is able to be used as voice input by other keyboards

@dessalines
Copy link

This would be wonderful to have. I've compared Vosk with Futo and Futo is nearly always better.

@domportera
Copy link

agreed - i can see myself using this very often with FUTO voice

@andsofine
Copy link

related to #169

@alexispurslane
Copy link

This would be hugely useful to me as well!

@Alexanderkocha
Copy link

I use Google's STT myself and I would love to be able to use it in this project as well.

@Stypox Stypox added enhancement New feature or request tts&stt Speech-to-text, text-to-speech and wakeword requests or bugs, including Vosk labels Feb 26, 2025
@Stypox
Copy link
Owner

Stypox commented Feb 26, 2025

Please let me know if #294 works. This PR implements one possible way to interact with external STT services, i.e. by opening the STT popup offered by another application (e.g. whisperIME: woheller69/whisperIME#53 (comment) ). So the PR only fixes one possible interpretation of this issue, while the other interpretation is using RecognizerService which would make for a far more seamless implementation.

Edit: I tested this with FUTO Voice input and it works quite well!

Screen_recording_20250227_005009.webm

@Alexanderkocha
Copy link

Sorry to reply after a month after this comment, however I am glad to be able to say it is working for me.

However one problem is that the region of STT defaults to my local setting and cannot be overwritten. Unfortunately, Google's STT does not work offline for my device's language.

I am not sure how keyboards like Microsoft SwiftKey get around this, but they seem to invoke Google STT in English by default.

So to summarise, it works, however due to my setup, I'm not able to make use of it "privately".

Many thanks for this PR, it's wonderful to see this project progress.

@Stypox
Copy link
Owner

Stypox commented Mar 30, 2025

Thanks for the report! This is caused by #294 (comment)

I tested with "speech recognition and synthesis from google", which should be the Google STT you are referring to, and it chooses the correct language now.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request tts&stt Speech-to-text, text-to-speech and wakeword requests or bugs, including Vosk
Projects
None yet
Development

No branches or pull requests

9 participants