In November last year, I presented a tutorial on waveform-based music processing with deep learning at ISMIR 2019. The tutorial covered music classification, source separation, and music generation in the waveform domain. The event attracted 450 attendees, providing a unique experience compared to larger machine learning conferences. I decided to write a blog post based on the tutorial content since it was well-received. The post covers motivation for waveform domain modeling, generative models, likelihood-based and adversarial models of raw music audio, and more. Waveform modeling is emphasized due to its ability to capture sound intricacies that other representations might miss.
https://sander.ai/2020/03/24/audio-generation.html