Dancing YMCA with Delayed Sensory Feedback

R.A. Løvlid and P. Öztürk (Norway)


Neural Networks, Feedforward Control, Internal Models


Lack of sensory feedback or delay in feedback has been shown to have detrimental effects on cognition and action in humans and in artifacts. Despite the adverse effects of delay, humans manage to generate smooth and timely movements. This has been explained by the existence of predictive models in the brain. In this paper we investigate the possible role of a predictive model that anticipates the consequences of the motor command to be issued to the actuator (e.g. arm). The paper presents two architectures, one with and the other without predictive components, and compares their performances in dancing to the song ‘YMCA’. The architecture including the predictive model has been trained in three different ways to uncover the possible effects of the training method on the movement performance. The results confirm the role of prediction in the movement control.

Important Links:

Go Back