Alisa Liu, Alex Fang, Gaetan Hadjeres, Bryan Pardo
Symbolic music generation uses machine learning to produce music in a symbolic form, such as the Musical Instrument Digital Interface (MIDI) format. Generating music in a symbolic format has the advantages of being both interpretable (e.g., as pitch, duration, and loudness values) and editable in standard digital audio workstations (DAWs).
Symbolic music generation also opens up the possibililty for users to condition the generation of music on encodings of both high-level semantic ideas (e.g., genre and mood) as well as low-level conditioning on existing musical content (e.g., a first verse or bassline).
[pdf] A. Liu, A. Fang, G. Hadjeres, P. Seetharaman, and B. Pardo, “Incorporating Music Knowledge in Continual Dataset Augmentation for Music Generation,” in Machine Learning for Media Discovery Workshop at the 37th International Conference on Machine Learning (ICML), 2020.
[pdf] A. Fang, A. Liu, P. Seetharaman, and B. Pardo, “Bach or Mock? A Grading Function for Chorales in the Style of J.S. Bach,” in Machine Learning for Media Discovery Workshop at the 37th International Conference on Machine Learning (ICML), 2020.