Our long-term goal is to reverse-engineer the human brain in order to understand how a system can construct the complex phenomenon that we call vision.
Many approaches to vision focus on the static aspects of vision and analyze how the three-dimensional structure of the world is estimated from the two-dimensional images on the retina. However, due to the movements of the observer, movements of the eyes, and the movements of objects, it is clear that natural vision is highly dynamic, as emphasized by various researchers such as Joseph Ternus, Gunnar Johansson, and James J. Gibson. The projection of the three-dimensional world on our retinae (proximal stimulus) undergoes complex real-time changes that are dependent on both the properties of our environment (distal stimulus) and our own movements.
REFERENCE-FRAMES AND THE METRIC OF VISUAL REPRESENTATIONS
Thus, the visual system needs to select in real-time appropriate reference-frames and metrics in order to disentangle the properties of the environment from those that result from our own actions.
Retinotopy, the initial representation in the visual system: The optics of the eyes map the three-dimensional environment into two-dimensional images on the retina. These two-dimensional representations, known as retinotopy, are preserved in early visual areas in the cortex.
EXAMPLE OF RETINOTOPY: Spatially neighboring stimuli are mapped on neighboring regions in early visual cortex. Top panels: Stimuli viewed by the observer; bottom panels: Areas activated by each stimulus (color coded) in early visual cortex.