Using Machine Learning for Stem Separation

This talk explores how modern machine learning techniques have advanced automatic audio stem separation – the process of isolating instruments from mixed recordings. We’ll look at the challenges involved in source separation, how recent AI developments have made high-quality separation possible, real-world applications in music production, and the technology behind AudioStrip.