Visual stability

From A conversation about the brain
Jump to: navigation, search

Visual stability is often seen as a big problem to solve, but it is not clear that it is so problematic. Essentially every movement we make comes with a prediction of the sensory consequences of that movement (e.g. gamma signals sent to muscle spindles) and movements that change the retinal image (e.g. saccades or head movements) are no exception. The debate might then turn to the source of this prediction. One possibility is the ‘computer graphics approach’ in which knowledge of the future camera/eye position and of the 3D structure of the scene can be combined to generate a predicted image. The framework described in these pages favours an approach based instead on stored sensory contexts.

Evidence about shifting receptive fields [1] has had a strong influence on the literature relating to visual stability. It has contributed to the idea that it would be helpful to transfer a large amount of data about the visual direction of objects into the current retinotopic frame[2]. It is not clear why this would be a helpful thing to do nor how it would work. One issue, for example, is how far back in time the data transfer goes (or should go): just visual information from the previous fixation? Does this include, recursively, all the information from earlier fixations? If not, what is excluded? Is all this data transfer necessary?

An alternative view, consistent with the framework described here, is that retinotopic information is ‘painted’ onto a ‘canvas’. As the eyes dart around the scene, information is added to a store or extracted from the scene to be used in controlling the task at hand. Of course, this begs the question about what makes up the ‘canvas’, but that is the central issue addressed in the section on 3D vision and a base representation. The important point is that the canvas must be stable not only when the eyes rotate (saccade) but also when the head translates, including translations by large distances.

A thoughtful discussion of the psychophysical and neurophysiological data that has been related to the issue of visual stability is by Cavanagh and colleagues[3].


  1. Duhamel, J.R., Colby, C. L., & Goldberg, M. E. (1992). The updating of the representation of visual space in parietal cortex by intended eye movements. Science, 255(5040), 90-92.
  2. Melcher, D. (2011). Visual stability. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1564), 468-475.
  3. Cavanagh, P., Hunt, A. R., Afraz, A., & Rolfs, M. (2010). Visual stability based on remapping of attention pointers. Trends in cognitive sciences, 14(4), 147-153.