Spleeter
2019-11-04 18:11:59.167645+00 by Dan Lyke 0 comments
Mind-blowing machine learning technology of the moment: Spleeter has TensorFlow models to split audio tracks. I've been playing with the "remove the voice track" model, and it's blowing my mind.
Related: If anyone has ideas for taking the vocal track and turning it into something that sounds more instrumental or synth-ish so that it doesn't compete with the vocals, I'd love to use this for automating square dance music creation.