Individual Connection Control in Central Pattern Generator for Music

The central pattern generators constructed from oscillating motor neurons are common biological phenomena. Central pattern generators are able to generate large number of different patterns, ranging from small periodic patterns to chaotic ones [Matsuoka, 1987]. Earlier the simulated oscillating neural systems have been used in rhythm perception research [Ohya, 1994] and musical pattern generation [Laine 2000]. Another area where simulated motor neurons have been used is the locomotion of legged robots [Cruse 1998]. The central pattern generator model consists of several model neuron objects which are connected with mutual inhibitory links. Inhibiting connection controls the activity increase of simple increase-and-fire type neurons. The inhibition is linked to other neurons without any delay. Different architectures (connection setups) affect the general style or behaviour of the connection. In earlier versions of the MINN-networks for music [Laine 1997, Laine 2000] 'neurons sensitivity' to all connections could be adjusted. In current system it is possible to adjust each connection separately, thus emphasizing the importance of each connection, rather than neuron, in the pattern generating network. Individual connection control is implemented by representing each connection as an instance of 'connection' class. One advantage of this method is that it makes possible the creation of 'handles' i.e. one or few pointers to important connections. Adjustment of these handles can be used to change the 'rhythmic gait' for example. Early experiments with the system indicate that handle-type control of pattern generator facilitates the preserving the basic style of rhythmic pattern yet allowing the adjustment of the pattern. Also the possibility of have several different connection strengths to each neuron facilitates more subtle control of the network. Similarly to earlier experiments the output of the pattern generator is mapped to musical parameters (volume/frequency or MIDI-values) for listening. Examples of the generated output will be played during the presentation.