3 ms·
> Give us some filters like "vocal fry", "perplexed", "angry", "wonder" etc and then we'll really have something here Neural networks can already do style tran
by Florin_Andrei 10y ago
> Give us some filters like "vocal fry", "perplexed", "angry", "wonder" etc and then we'll really have something here
Neural networks can already do style transfer for images. Seems like it should be doable for voice, too.
- augustt 10y agoI wonder if you could get interesting results by applying neural style transfer to spectrograms.