Live Subtitle Generation

[sooswastaken]

What you want IINA to do:

I would love a feature where you could have subtitles be autogenerated. Chrome has a feature called Live Caption, and I presume it scans the audio being played and generates subtitles.

What IINA does currently:

Doesn't have Live Caption

Why you think this should be added:

If subtitles aren't available for a certain video, people could use this feature to generate them. It would be really helpful for people with disabilities.

Examples of other projects that have something similar:

Google Chrome

[low-batt]

An interesting and challenging feature proposal.

This would require finding an embeddable, on device, open source, speech to text library and integrating with it. Something along the lines of DeepSpeech.

[low-batt]

@svobs Mentioned Whisper as an example of another open source speech to text application.

[sxflynn]

Here is a native implementation https://github.com/ggerganov/whisper.cpp

[low-batt]

Thanks for posting that link!

IINA is likely to follow recommendations in the Accessibility section of Apple's Human Interface Guidelines. From a quick look at the HIG I did not spot anything regarding this, but my guess is that Apple will expect applications to rely upon Live Captions for this once that feature is out of beta.