Explore the application of deep learning to music analysis and generation in this 44-minute talk from the EuroPython 2018 Conference. Delve into techniques for detecting instruments, identifying musical elements, and isolating individual instruments in polyphonic compositions. Learn how to create a simple synthesizer using NumPy and generate an unlimited dataset of musical improvisations. Discover the process of adapting deep learning models to handle increasing musical complexity, from simple melodies to multi-instrument arrangements. Gain valuable insights into effective deep neural network structures for music processing and generation.
Overview
Syllabus
Marcel Raas - Change music in two epochs
Taught by
EuroPython Conference