Vision Sciences: Depth perception in complex scenes

When we view a 3D scene in the cinema or on a 3D TV, we often obtain and compelling and impressive sensation of depth, object shape, and the space between objects. This is possible because the tiny differences between right and left eye views of the world can be interpreted by our visual systems so that we perceive depth. The complex processing required to achieve this is not yet fully understood. This project aims to use existing models of the early stages of binocular visual processing to predict and explain when binocular information will enhance depth perception for complex 3D scenes. The project will involve human experimental techniques and computational modelling of the visual system. The student will be based in the Vision Lab at the University of St. Andrews and will be part of an interdisciplinary team working at the interface between perceptual psychology, neuroscience and computational vision.

The studentship is part of a larger project in the group that aims to develop a detailed understanding of the utility of binocular depth information in human visual processing. It will provide crucial core knowledge about visual processing of the 3-D world by the brain, that will feed into improving the realism of rendered 3-D images and movies, and the accuracy of computer vision algorithms. Team members on this project come from diverse disciplines (Computer Science, Physics, Psychology). The successful candidate will be willing and able to embrace our interdisciplinary approach, and will have a 1st Class or 2:1 (or equivalent) first degree in computer science, a physical science, engineering, mathematics, psychology, neuroscience or a related discipline. Good quantitative skills are essential, as is mathematics to at least A-level standard, or equivalent.