out of distribution

Black music powered every significant musical movement.

Every.

thank u ben i tried i practiced 🙂‍↕️🙏❤️

Great work from @hugofloresgarcia.bsky.social’s internship at Adobe - turn your voice into basically anything!

Sketch2Sound generates audio from time-varying control signals (loudness, brightness, pitch) and text prompts, using sonic imitations; it uses random median filters on control signals during training for flexible temporal specificity.

here's a guitar -> sfx example!

Sketch2Sound can be implemented on top of any text-to-audio DiT and requires 40k steps of fine-tuning and a single linear layer per control!

in collaboration w/
@urinieto.bsky.social, @justinsalamon.bsky.social, #bryanpardo and the legendary @pseeth.bsky.social!

new paper! 🗣️Sketch2Sound💥

Sketch2Sound can create sounds from sonic imitations (i.e., a vocal imitation or a reference sound) via interpretable, time-varying control signals.

paper: arxiv.org/abs/2412.08550
web: hugofloresgarcia.art/sketch2sound

beep!! 🤠