Summary

Gaze control involves eyes, head, and body movements and is guided by mainly three types of information: visual, vestibular, and proprioceptive. Appropriate gaze control is a basis for actions such as reaching, grasping, eating, and manipulation, all of which develop during the first year of life. The development of gaze control is about how young infants gain access to these different kinds of information, how they come to use them, and how they come to coordinate head and eyes to accomplish it. This control develops during the first few weeks of life. A major challenge for the gaze controlling system is how gaze is stabilized on a moving target to keep vision clear, including during self-motion or the compensation of other sudden movements. Furthermore, the tracking has to be timed relative to the object motion. This requires prediction, which is a part of smooth pursuit that emerges at around six weeks and is in full function at three months. The smooth eye and head movements must add up in time and space to the object motion. Then the vestibular and visual neural signals must be properly added. Catch-up saccades compensate when the smooth pursuit is insufficient. In other situations, saccades shift the gaze between objects or situations. Moreover, if a moving object temporarily disappears out of view, one or several saccades predictively recapture the object at the reappearance position (four months). The complex and fast development of gaze has inspired the design of robotic vision (iCub) through processes similar to human development, thus increasing the robot’s flexibility and learning abilities