Pre-Publication V3. The application of machine learning models algorithms that have been trained on an data set of existing catalog, library, or other source of music to generate derivative compositions from a given input prompt.
Deliberation Summary:
(1) Suggest describing the systems as models rather than as algorithms, and noted that the list of sources of music was redundant and could be reduced to a data set of existing music. (2) Because referring to a prompt implies text alone, a general reference to input covers audio as well.
Pre-Publication V2. Musical Generation
The application of machine learning algorithms that have been trained on an existing catalog, library, or other source of music to generate derivative new compositions from a prompt.
Deliberation Summary:
(1) Consider whether compositions produced should be described as new; suggest using derivative as the more accurate term. (2) Edits to expand on the description of the source material.
Pre-Publication V1. The application of machine learning algorithms that have been trained on an existing catalog of music to generate new compositions from a prompt.