Real-time jam-session support system
This work addresses the problem of automated musical support for musicians during jam sessions, but it appears incremental as it builds on existing Markov model techniques.
The authors tackled real-time chord accompaniment for improvised music by predicting the next chord using a combination of Hidden Markov Models and Variable Order Markov Models, achieving results evaluated through both objective accuracy and subjective questionnaires.
We propose a method for the problem of real time chord accompaniment of improvised music. Our implementation can learn an underlying structure of the musical performance and predict next chord. The system uses Hidden Markov Model to find the most probable chord sequence for the played melody and then a Variable Order Markov Model is used to a) learn the structure (if any) and b) predict next chord. We implemented our system in Java and MAX/Msp and compared and evaluated using objective (prediction accuracy) and subjective (questionnaire) evaluation methods.