Robust, hybrid algorithms in AI-aided automatic music production
Pawe� Skrzy�ski, Joanna Kwiecień, Marek Pluta, Andrzej D�browski, Bart�omiej Szadkowski · AGH University of Science and Technology Press eBooks · 2022
Contemporary music market is vastly dominated by streaming services.Growing popularity of streamed music consumption enables each music genre to find its niche.Some genres are defined well enough in terms of musical form and structure to carry out attempts at making the process of their production fully automated.This work presents a selection of algorithms designed for automatic music production.The algorithms are implemented in a fully functional production system that uses a hybrid approach, combining robustness and controllability of classic solutions with ability to generate and evaluate complex structures inherent for AI-based methods.So far, the approach has been applied to produce relaxation, dance, and electronic musicgenres that are popular in streaming services, and well suited for automatic production.