Multiple Timescales Recurrent Neural Network (MTRNN) Model

This page introduces two representative studies of utilizing MTRNN model to achieve cognitive behaviors of humanoid robots.

(1) This study examines how functional hierarchy can self-organize through sensory-motor interactions, without assuming predefined level-structured functions. A humanoid robot was implemented with so-called the multipe timescales recurrent neural network (MTRNN). The MTRNN consists of the fast neurons part and the slow neurons one which are interconnected each other within a single network.
The results of the robot learning experiments showed that functional hierarchy emerges with accompanying a compositional structure such that the continuous sensory-motor flow is segmented into reusable behavior primitives in the fast neurons part and those primitives are integrated into specified goal-directed actions in the slow neuron part.
A movie of this robotics experiment can be seen at this page.

Y. Yamashita and J. Tani: "Emergence of functional hierarchy in a multiple timescale neural network model: a humanoid robot experiment", PLoS Computational Biology, Vol.4, Issue.11, e1000220, 2008.

(2) This study suggests that deterministic chaos self-organized in cortical dynamics could be responsible for the generation of spontaneous action sequences. We conducted experiments of tutoring a humanoid robot for stochastic transition sequences of primitive actions by utilizing MTRNN. The experimental results showed that the robot successfully learned to imitate tutored behavioral sequence patterns by extracting the underlying transition probability among primitive actions. An analysis revealed that a set of primitive action patterns was embedded in the fast dynamics part and the chaotic dynamics of spontaneously sequencing these action primitive patterns was structured in the slow dynamics part, provided that the timescale was adequately set for each part. By making analogy of the model to the known cortical hierarchy for action generation, it is speculated that the slow dynamics part of self-organizing chaos might correspond to the prefrontal cortex that is known for its flexible production of actions. The study also sheds a ray of light on the complementary relation between learning of stochastic processes and that of deterministic ones.
A movie of this robotics experiment can be seen at this page.

J. Namikawa, R. Nishimoto, J. Tani: "A neurodynamic account of spontaneous behaviour", PLoS Computational Biology, Vol.7, Issue.10, e1002221, 2011.

Dr. Jun Tani, Riken Brain Science Institute