2D to 3: the visual representation of space in moving observers. (360G-Wellcome-086526_A_08_Z)

As we move through the world, we acquire visual information that allows us to build up a stable representation of the scene. The aim of this project is to challenge current models of this process and inform the development of new theoretical accounts. Our high fidelity virtual reality laboratory, which relies on a close collaboration with engineering colleagues, provides a unique opportunity to test hypotheses about spatial representation in moving observers using controlled psychophysical techn iques. The current 'textbook' model of human spatial representation is Cartesian, in which there is a 1:1 mapping between physical space and its internal representation. We will test the following predictions of this model: (i) transitivity of depth judgements; (ii) independence of visual space distortions across tasks; (iii) covariance between size and distance judgements when the reliability of one cue is varied; (iv) consistency in the visual direction of objects with observer movement. We will also collaborate with colleagues in Robotics at Oxford to test alternative, non-Cartesian, view-based approaches to `homing'. A long term goal is to develop view-based models of peri-personal space and to test these in the same way that we propose here for the Cartesian approach.