Grant: Operating Grant

Our rich and effortless perception of the complex mixtures of objects and surfaces comprising the everyday visual world is one of the most remarkable achievements of the human brain. Immediate and accurate parsing of this rich sensory array is crucial in order for us to perform everyday tasks such as object recognition, spatial cognition and visuomotor control. Such high-level visual operations can only be achieved on the basis of meaningful low-level neural representations of gradients and discontinuities of simple visual features such as changes in luminance, contrast, texture, or motion. These neural signals can enable critical basic operations such as detecting figure-ground boundaries, or disambiguating whether contours arise from changes in surface reflectance, illumination, or occlusion boundaries. These low-level processes necessarily precede higher-level visual functions such as object recognition, and are fundamentally important to normal visual perception. The outcomes of this work may enable our understanding of anomalies of the visual system in aging and in clinical pathologies, and the design of tests to assess them. This research will focus on evaluating the idea that gradient or boundary information from luminance and from non-luminance (contrast, texture, motion) cues are initially detected by separate mechanisms, and then combined in different ways to solve different computational problems. The proposed projects will utilize single unit and local population electrophysiology with multi-electrodes, optical imaging, and system identification using natural images, exploiting modern methods of spike waveform classification, time-spectral histograms, and current source density analysis. This research is novel and significant because it attacks the problem of how simple visual features are assembled to provide a basis for perception in complex everyday scenes.