Nao Tokui

Prototyping an M4L device that “translates” rhythm patterns into corresponding basslines. The device uses a seq-to-seq model, normally used for language translation tasks, trained on house music MIDI files containing both drum patterns and bass lines.

The basic assumption behind this device is the following: A drum pattern and its corresponding bass line, they are speaking the same rhythmical concept in two different languages.

The device and source code are available on our GitHub page.

Reference:  Behzad Haki, & Jorda, S. (2019). A Bassline Generation System Based on Sequence-to-Sequence Learning. Proceedings of the International Conference on New Interfaces for Musical Expression, 204–209. http://www.nime.org/proceedings/2019/nime2019_040.pdf