The OMax project is based at Ircam with collaborations from UCSD, CNSMDP and EHESS. OMax is a technology that allows a virtual musical agent to learn straight from the musicians playing live and to develop its own musical discourse in dialog with the musicians based on stylistic and timbral imitation.
SoMax is a new OMax research thread developed by Laurent Bonasse-Gahot in the context of the IRCM ANR Project SOR2.
Based on corpus machine learning it generates improvised accompaniment to any kind of inputs (midi, audio, monodic, polyphonic) in the form of augmented rendering (harmony, voicings, orchestrations)
-on "SoMax" SoMax interacts with the saxophonists and provides real time improvised harmonic augmentation of the input.
OMax the virtual improviser is an Ircam / RepMus team project, with contributinons by G. Assayag, G. Bloch, M. Chemillier, B. Lévy, S. Dubnov, L. Bonnasse-Gahot