Is there a method of extracting a prompt from a given audio sample?

facebookresearch / audiocraft

Audiocraft is a library for audio processing and generation with deep learning. It features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.

MIT License

20.5k stars 2.06k forks source link

Is there a method of extracting a prompt from a given audio sample? #249

Closed slooi closed 12 months ago

slooi commented 1 year ago

Title. This would allow us to find what elements we like in existing audio samples and use them in our music generations

carlthome commented 12 months ago

Guess you could view that inverse problem as traditional music auto-tagging but it's not gonna retrieve the prompt itself and rather just predict keywords: https://paperswithcode.com/task/music-auto-tagging

adefossez commented 12 months ago

This is not really possible and is an open research question !