Eye movements in depth to visual illusions

Abstract

We perceive the three-dimensional (3D) environment that surrounds us with deceptive effortlessness. In fact, we are far from comprehending how the visual system provides us with this stable perception of the (3D) world around us. This thesis will focus on the interplay between visual perception of depth and its closely related action system, eye movements in depth.

The human visual system is comprised of a sensory (input) and an output (motor) system. Processed information from the sensory system can result in two explicit measurable response types: conscious visual perception and ocular motor behavior. It is still a matter of debate whether conscious visual perception and action (including hand- and arm-movements) use the same information or whether the visual system has separate channels processing information for perception and action. In this thesis, we study (1) if separate channels, one for eye movements and one for conscious visual perception, indeed exist, and (2) if so, if there is a direct input from the perceptual pathway to the motor pathway. Assuming that either eye movements and conscious visual perception are based on information from a common source (a negative answer to issue 1) or perception can directly influence, or guide, eye movements (an affirmative answer to research question 2), (eye) movements reflect our conscious visual perception. If so, eye movements could provide us with an alternative method to probe our conscious visual perception, making explicit perceptual reports superfluous.

In this thesis we focus on depth perception and the two types of eye movements that are closest related to depth perception, namely vergence (an eye movement that gets a certain depth plane into focus) and saccades (a rapid eye movement to change gaze direction). Over the last 20 years it has been shown that depth perception is based on a weighted combination of depth cues available such as linear perspective, occlusion and binocular disparity. How eye movements are planned, however, is still unclear. Several studies have reported that eye movements are, to varying degrees, correlated with perception and thus concluded that perception guides eye movements. However, in most of these studies depth perception was correlated to the depth cues and a clear distinction between cues and perception could not be made.

A way to make a dissociation between cues and perception, is to make use of depth reversal illusions: stimuli that can induce multiple equally likely depth interpretations while the stimulus cues remain the same. That means that perception can alternate, while cues remain constant leading to a dissociation between perception and cues.

In several different studies we show that in the case of vergence, eye movements are planned based on depth cues (mainly disparity) and are uncorrelated to perception. In the case of saccades, we show that the direction of saccades is highly correlated to perception, but seems to be subserved by a separate system combining cues using very similar weights as for perception.