⌘K
0% complete
0 reference numerals
~1 pages
Jurisdiction
United States (USPTO)
Type
Design
Entity Size
Small
Technology
A Music Creation Engine.
A method for generating a musical composition using a machine learning model trained on a multi-genre dataset comprising receiving, by a computing system, a user-defined prompt specifying one or more musical attributes; selecting, by the computing system, a machine learning model trained on the multi-genre dataset based on the user-defined prompt; generating, by the computing system using the machine learning model, a song structure defining a plurality of sections; generating, by the computing system using the machine learning model, instrumentation for the plurality of sections; generating, by the computing system using the machine learning model, a harmonic progression for at least a portion of the plurality of sections; generating, by the computing system using the machine learning model, production parameters comprising one or more of tempo, key, dynamics, timbre, effects, and mixing settings; rendering, by the computing system, an arranged musical composition comprising the song structure, the instrumentation, the harmonic progression, and the production parameters; and outputting, by the computing system, the arranged musical composition in a format configured for commercial distribution and adaptive playback environments
The method for generating a musical composition using a machine learning model trained on a multi-genre dataset comprising wherein receiving the user-defined prompt comprises receiving, via a user interface, at least one of a genre identifier, a mood identifier, an instrumentation constraint, a target duration, a target loudness, and a reference audio clip, and wherein selecting the machine learning model comprises selecting a plurality of genre-conditioned model parameters based on the at least one of the genre identifier and the reference audio clip
The method for generating a musical composition using a machine learning model trained on a multi-genre dataset comprising further comprising optimizing, by the computing system, the arranged musical composition for adaptive playback environments by generating a plurality of stems corresponding to at least a melody stem, a harmony stem, a bass stem, a percussion stem, and an effects stem, and encoding, in association with the plurality of stems, playback metadata operable to control dynamic mixing of the plurality of stems based on a playback context