Compensation of oculomotor delays in the visual system's network
- a poster @ the “Complex Networks: from theory to interdisciplinary applications” conference which will take place on July 11-13, 2016 in Marseilles, France, as a satellite meeting to Statphys26 (July 18-22, Lyon, France).
get the poster
- Laurent U. Perrinet, Rick A. Adams, Karl Friston. Compensation of oculomotor delays in the visual system's network., URL . In Complex Networks: from theory to interdisciplinary applications., pages paper 61. 2016 abstractWe consider the problem of sensorimotor delays in the optimal control of movement under un-certainty. Speciﬁcally, we consider axonal conduction delays in the visuo-oculomotor loop andtheir implications for active inference. Active inference uses a generalisation of Kalman ﬁlter-ing to provide Bayes optimal estimates of hidden states and action in generalised coordinatesof motion. Representing hidden states in generalised coordinates provides a simple means ofcompensating for both sensory and oculomotor delays. This compensation is illustrated us-ing neuronal simulations of oculomotor following responses with and without compensation.We then consider an extension of the generative model that produces ocular followingto simulate smooth pursuit eye movements — in which the system believes both the target andits centre of gaze are attracted by a (ﬁctive) point moving in the visual ﬁeld. Finally, the generative model is equipped with a hierarchical structure, so that it can register and rememberunseen (occluded) trajectories and emit anticipatory responses. These simulations speak to astraightforward and neurobiologically plausible solution to the generic problem of integratinginformation from different sources with different temporal delays and the particular difficultiesencountered when a system — like the oculomotor system — tries to control its environmentwith delayed signals..
All material (c) L. Perrinet. Please check the copyright notice.
This work was supported by European Union project Number FP7-269921, "BrainScales".