Re:I made a free transcription service powered by Whisper AI
by ivalm on Saturday November 19, 2022 @12:07AM
(#33667140)
Could one train an interpolation layer (eg take a bunch of 16k audio, down sample to 8k, train 8k->16k upsampler)? Or better yet (but more expensive), take whisper, freeze it, and train the upsampler on whisper’s loss.
[ Parent
| on HN ]
Re:I made a free transcription service powered by Whisper AI
by kkielhofner on Saturday November 19, 2022 @12:16AM
(#33667191)
"I understand some of those words."
Hah, in all seriousness I'm more of a practitioner in this space. If this was something I absolutely needed to get done who knows where it would have went. For a little side hacking project once I encountered these issues I moved on - back in the day expectations were lower for telephony and the 8 kHz aspire models and kaldi were adequate to get that "real work" done.
Re:I made a free transcription service powered by Whisper AI
by andrew3726 on Saturday November 19, 2022 @05:15AM
(#33668845)
Sure, that's called audio super resolution, there's a few papers/projects doing that. Haven't really seen models which are robust and have good generalization though.