Grants and Contributions:

Grant or Award spanning more than one fiscal year. (2017-2018 to 2022-2023)

The goal of this ongoing research program is to develop a rigorous understanding of the interaction between gaze shifts and visual perception. Behavioral experiments conducted over the last 100 years have noted a wide range of changes in visual perception around the time of fast eye movements called saccades. These include illusory changes in stimulus position, timing, and even shape. A similar range of perceptual effects has been found for slower eye movements known as smooth pursuit. Together these effects reveal a clear influence of oculomotor signals on visual perception. Yet they have typically been studied under rather artificial conditions, in which subjects are instructed to make a single eye movement, while an isolated visual stimulus is presented.

We hypothesize that these perceptual illusions are a side effect of neural mechanisms that actually improve visual function under more natural conditions. The main goal of this proposal is to examine this idea in quantitative detail.

Our model for understanding the interaction of vision and eye movements is the visual cortex of the macaque monkey. In particular, we study neurons in areas V4, IT, MT, and MST, which have are involved in integrating visual information with oculomotor signals related to saccades and smooth pursuit. Previous work from my lab and others has demonstrated profound changes in visual processing in these areas around the time of eye movements. However, there has been little attempt to formally link these neural changes to the perceptual changes that are seen in human subjects. Moreover, the functional benefit of these changes is unknown: How does the specific pattern of oculomotor influences on visual processing improve our ability to localize and recognize objects? There has been little attempt to answer this question, and some have argued that vision is not improved by oculomotor inputs at all.

The proposed research will address this issue by developing a novel framework for interpreting visual processing during eye movements. This will be based on multi-electrode recordings during eye movements and during steady fixation. Together these data will be used to characterize mathematically the coding of stimulus information in neuronal populations in the visual cortex. Importantly, this framework leads to specific predictions about the role of oculomotor influences in natural vision: for localizing objects in space, for identifying them, and for estimating their velocity. We will test these predictions by recording from the same neurons while the animals freely view natural scenes. This framework will then be extended to the study of combined eye-head movements, to determine whether the computational advantages of visual-oculomotor integration translate to eye-head gaze shifts as well. Together these results will provide insight into the functional interaction between visual and oculomotor signals in the primate brain.