Hi

On Wed, Jul 09, 2025 at 11:24:26PM +0800, Zhao Zhili wrote:
> 
> > On Jul 9, 2025, at 15:23, Vittorio Palmisano <vpalmisano-at-gmail.com@ffmpeg.org> wrote:
> > 
> > It adds a new audio filter for running audio transcriptions with the whisper model.
> > Documentation and examples are included into the patch.
> 
> The patch doesn’t following ffmpeg coding style.
> 
> Setting aside the coding style issues, I have a few concerns.
> 

> There are DNN support with three backends in FFmpeg (libavfilter/dnn_interface.h), which
> are supposed to be robust extensibility.

If someone implements some high quality speech recognition more natively
that surely would be cool. Our own model would be cool too, ...
or supporting the whisper models natively, all that would be very cool


> I guess incorporating Whisper natively into our DNN
> architecture can be difficult, making wrapper another library more feasible than direct
> integration.

yes, i think so too

thx

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

No human being will ever know the Truth, for even if they happen to say it
by chance, they would not even known they had done so. -- Xenophanes