论文信息 - The Role of Internal Oscillators for the One-Shot Learning of Complex Temporal Sequences

The Role of Internal Oscillators for the One-Shot Learning of Complex Temporal Sequences

We present an artificial neural network used to learn online complex temporal sequences of gestures to a robot. The system is based on a simple temporal sequences learning architecture, neurobiological inspired model using some of the properties of the cerebellum and the hippocampus, plus a diversity generator composed of CTRNN oscillators. The use of oscillators allows to remove the ambiguity of complex sequences. The associations with oscillators allow to build an internal state to disambiguate the observable state. To understand the effect of this learning mechanism, we compare the performance of (i) our model with (ii) simple sequence learning model and with (iii) the simple sequence learning model plus a competitive mechanism between inputs and oscillators. Finally, we present an experiment showing a AIBO robot, which learns and reproduces a sequence of gestures.

[1] Mathias Quoy,et al. Resonant spatiotemporal learning in large random recurrent networks , 2002, Biological Cybernetics.

[2] Randall D. Beer,et al. Sequential Behavior and Learning in Evolved Dynamical Neural Networks , 1994, Adapt. Behav..

[3] A. Billard,et al. A Biologically-Inspired Model of Reaching Movements , 2006 .

[4] Philippe Gaussier,et al. Learning Invariant Sensorimotor Behaviors: A Developmental Approach to Imitation Mechanisms , 2004, Adapt. Behav..

[5] Yves Coiton,et al. A neural network model for temporal sequence learning and motor programming , 1994, Neural Networks.

[6] Sorin Moga,et al. From Perception-Action loops to imitation processes , 1998 .

[7] Auke Jan Ijspeert,et al. A neuromechanical investigation of salamander locomotion , 2000 .

[8] Aude Billard,et al. Learning of Gestures by Imitation in a Humanoid Robot , 2007 .

[9] Jun Nakanishi,et al. Nonlinear dynamical systems for imitation with humanoid robots , 2001 .

[10] Hugues Bersini,et al. Learning Cycles brings Chaos in Continuous Hopfield Networks , 2005 .

[11] Sorin Moga,et al. From Perception-Action Loops to Imitation Processes: A Bottom-Up Approach of Learning by Imitation , 1998, Appl. Artif. Intell..

[12] Daniel Bullock,et al. Neural dynamics of planned arm movements: emergent invariants and speed-accuracy properties during trajectory formation , 1988 .