ETD system

Electronic theses and dissertations repository


Tesi etd-06062012-183701

Thesis type
Tesi di dottorato di ricerca
Gaze control modelling and robotic implementation
Settore scientifico disciplinare
Corso di studi
relatore Prof. Dario, Paolo
correlatore Prof. Berthoz, Alain
tutor Prof.ssa Laschi, Cecilia
Parole chiave
  • optimal control
  • internal models
  • human locomotion
  • gaze control
  • eye movements
  • trajectory planning
Data inizio appello
Riassunto analitico
Although we have the impression that we can process the entire visual field in a single fixation, in reality we would be unable to fully process the information outside of foveal vision if we were unable to move our eyes. Because of acuity limitations in the retina, eye movements are necessary for processing the details of the array. Our ability to discriminate fine detail drops off markedly outside of the fovea in the parafovea (extending out to about 5 degrees on either side of fixation) and in the periphery (everything beyond the parafovea). While we are reading or searching a visual array for a target or simply looking at a new scene, our eyes move every 200-350 ms. These eye movements serve to move the fovea (the high resolution part of the retina encompassing 2 degrees at the centre of the visual field) to an area of interest in order to process it in greater detail. During the actual eye movement (or<br>saccade), vision is suppressed and new information is acquired only during the fixation (the period of time when the eyes remain relatively still). While it is true that we can move our attention independently of where the eyes are fixated, it does not seem to be the case in everyday viewing. The separation between attention and fixation is often attained in very simple tasks; however, in tasks like reading, visual search, and scene perception, covert<br>attention and overt attention (the exact eye location) are tightly linked. Because eye movements are essentially motor movements, it takes time to plan and execute a saccade. In addition, the end-point is pre-selected before the beginning of the movement. There is considerable evidence that the nature of the task influences eye movements. Depending on the task, there is considerable variability both in terms of fixation durations and saccade lengths.<br>It is possible to outline five separate movement systems that put the fovea on a target and keep it there. Each of these movement systems shares the same effector pathway—the three bilateral groups of oculomotor neurons in the brain stem. These five systems include three that keep the fovea on a visual target in the environment and two that stabilize the eye during head movement. Saccadic eye movements shift the fovea rapidly to a visual target<br>in the periphery. Smooth pursuit movements keep the image of a moving target on the fovea. Vergence movements move the eyes in opposite directions so that the image is positioned on both foveae. Vestibulo-ocular movements hold images still on the retina during brief head movements and are driven by signals from the vestibular system. Optokinetic movements hold images during sustained head rotation and are driven by visual stimuli. All eye movements but vergence movements are conjugate: each eye moves the same amount in the same direction. Vergence movements are disconjugate: The eyes move in different directions and sometimes by different amounts. Finally, there are times that the eye must stay still in the orbit so that it can examine a stationary object. Thus, a sixth system, the fixation system, holds the eye still during intent gaze. This requires active suppression of eye movement. Vision is most accurate when the eyes are still. When we look at an object of interest a neural system of fixation actively prevents the eyes from moving. The fixation system is not as active when we are doing something that does not require vision, for example, mental arithmetic. Our eyes explore the<br>world in a series of active fixations connected by saccades. The purpose of the saccade is to move the eyes as quickly as possible. Saccades are highly stereotyped; they have a standard waveform with a single smooth increase and decrease of eye velocity. Saccades are extremely fast, occurring within a fraction of a second, at speeds up to 900°/s. Only the distance of the target<br>from the fovea determines the velocity of a saccadic eye movement. We can change the amplitude and direction of our saccades voluntarily but we cannot change their velocities. Ordinarily there is no time for visual feedback to modify the course of the saccade; corrections to the direction of movement are made in successive saccades. Only fatigue, drugs, or pathological states can slow saccades. Accurate saccades can be made not only to visual targets but also to sounds, tactile stimuli, memories of locations in space, and even verbal commands (“look left”). The smooth pursuit system keeps the image of a moving target on the fovea by calculating how fast the target is moving and moving the eyes accordingly. The system requires a moving stimulus in order<br>to calculate the proper eye velocity. Thus, a verbal command or an imagined stimulus cannot produce smooth pursuit. Smooth pursuit movements have a maximum velocity of about 100°/s, much slower than saccades. The saccadic and smooth pursuit systems have very different central control systems. A coherent integration of these different eye movements, together with the other<br>movements, essentially corresponds to a gating-like effect on the brain areas controlled. The gaze control can be seen in a system that decides which action should be enabled and which should be inhibited and in another that improves the action performance when it is executed. It follows that the underlying guiding principle of the gaze control is the kind of stimuli that are presented to the system, by linking therefore the task that is going to be executed.<br>This thesis aims at validating the strong relation between actions and gaze. In the first part a gaze controller has been studied and implemented in a robotic platform in order to understand the specific features of prediction and learning showed by the biological system. The eye movements integration opens the problem of the best action that should be selected when a new<br>stimuli is presented. The action selection problem is solved by the basal ganglia brain structures that react to the different salience values of the environment. In the second part of this work the gaze behaviour has been studied during a locomotion task. The final objective is to show how the different tasks, such as the locomotion task, imply the salience values that drives the gaze.