New top story on Hacker News: Show HN: Cloning a musical instrument from 16 seconds of audio

Show HN: Cloning a musical instrument from 16 seconds of audio
19 by abdljasser2 | 1 comments on Hacker News.
In 2020, Magenta released DDSP [1], a machine learning algorithm / python library which made it possible to generate good sounding instrument synthesizers from about 6-10 minutes of data. While working with DDSP for a project, we realised how it was actually quite hard to find 6-10 minute of clean recordings of monophonic instruments. In this project, we have combined the DDSP architecture with a domain adaptation technique from speech synthesis [2]. This domain adaptation technique works by pre-training our model on many different recordings from the Solos dataset [3] first and then fine-tuning parts of the model to the new recording. This allows us to produce decent sounding instrument synthesisers from as little as 16 seconds of target audio instead of 6-10 minutes. [1] https://ift.tt/cdPY8O9 [2] https://ift.tt/xUdJoz8 [3] https://ift.tt/aFO0Pvo We hope to publish a paper on the topic soon.

Comments

Popular posts from this blog

Northeastern US braces for foot of snow during first days of December

North Korea test fires two missiles month before deadline for US to respond on talks

Democratic debate winners and losers: Elizabeth Warren triumphs while Beto O'Rourke flounders