Automatic music accompaniment is particularly useful for exercises, rehearsals and personal enjoyment of ensemble music and one-hand piano performances. As musicians may make errors and want to correct them, or they may want to skip hard parts in the score, the system should allow errors as well as arbitrary repeats and skips. Detecting such repeats/skips, however, involves a large complexity of search for a player's score position in the entire score for every input event. Several efficient algorithms have been developed to cope with this problem under practical assumptions used in an online automatic accompaniment system named "Eurydice". In Eurydice for MIDI instruments, music performance is modeled by a hidden Markov model and maximum probability estimation is applied to the polyphonic MIDI input to yield an accompanying MIDI output (e.g., orchestra sound). Another version of Eurydice accepts monaural audio signal input and accompanies to it. Other issues such as treating ornaments, tempo estimation, and accompaniment algorithms are also discussed.

This content is only available via PDF.