Improvisation Based on Relaxation of Imitation Parameters by a Robotic Acoustic Musical Device

A new approach to musical improvisation based on controlled relaxation of imitation parameters by a robotic acoustic musical device, is presented in this paper. The presented RObotic Musical Instrument “ROMI”, is aimed at jointly playing two instruments that belong to two different classes of acoustic instruments and improvises while assisting others in an orchestral performance. ROMI’s intelligent control architecture also has the ability to provide player identification and performance training. In this paper we introduce the robotic device ROMI together with its control architecture, the Musical State Representation “MSR” and focus on parameter estimation for imitation of duo players by ROMI. The MSR we have developed for controlling ROMI is a feature extractor for learning to imitate human players in a duet. The control architecture has an automatic parameter estimation process that is usually employed for optimization of the imitation stage. Improvisation can be achieved by programming this process to function at non-optimal values, thereby sounding notes that are definitely different than the music piece being imitated. If this programming is not done without constraints then the resultant sound will deviate too much form the original music piece being imitated and can not be classified as an improvisation. The constraints of this programming are also introduced in this paper.