4 ms·
That depends on the model's training set. If the training set involves images of a lot of valid music, then probably it could generate "valid" sounding music.
by all2 3y ago
That depends on the model's training set. If the training set involves images of a lot of valid music, then probably it could generate "valid" sounding music.
Otherwise, it may be generating garbage. Can you post the image somewhere? I'd love to take a look.
- gus_massa 3y agoIt would be interesting that you take a look, but expect garbage. A few days ago someone posted AI generated poster of movies for children, and in one of them he studio was "Dixxar" (with a nice cursive D). I don't expect long text or music to make sense in an AI generated image for now.
- dukeofdoom 3y agoI don't know how to do this without doxxing myself, other wise I would.