Machine Pianist - MIDI Performance Augmentation

Using the power of machine learning to humanize automatic piano performances to the benefit of both composer and audience. Established under the hypothesis that the key "humanizing" components lacking in widely available online MIDI files are:

1. Variable velocity of notes when pressed
2. Subjective choice of pedal positions throughout pieces

Given an input MIDI file lacking such details, augment the file with key velocities and pedal positions to produce a more human-sounding performance for any arbitrary piano piece.

To use the publicly deployed model, please see the Machine Pianist website.

---

---

References

[1] Hawthorne, Curtis, et al. "Enabling factorized piano music modeling and generation with the MAESTRO dataset." arXiv preprint arXiv:1810.12247 (2018).