Observational Learning Based On Models Of Overlapping Pathways

Brain imaging studies in macaque monkeys have recently shown that the observation and execution of specific types of grasp actions activate the same regions in the parietal, primary motor and somatosensory lobes. In the present paper we consider how learning via observation can be implemented in an artificial agent based on the above overlapping pathway of activations. We demonstrate that the circuitry developed for action execution can be activated during observation, if the agent is able to perform action association, i.e. relate its own actions with the ones of the demonstrator. In addition, by designing the model to activate the same neural codes during execution and observation, we show how the agent can accomplish observational learning of novel objects.