Bayesian Decision Theory Application for Double-step Saccades

Abstract

Bayesian decision theory applied to the human saccadic eye movements elicited in double-step mode: primary and corrective saccade. Primary saccade performs large amplitude but not precise eye jump towards new target position and small amplitude corrective saccade brings line of sight precisely on the target. This behavior mach two-step Bayesian decision making process: first with large uncertainty and second, after getting additional information, – more precise. It explains how saccadic eye movements control system performs precise eye jump in the conditions of the noisy visual estimation of the new target position and eye jump errors made by eye globe muscles. Scatter of the landing places of the primary saccades was investigated theoretically using Bayesian statistics approach and experimentally. Ill. 5, bibl. 6 (in English; summaries in English, Russian and Lithuanian).