NEURAL CORRELATES OF MOVING BOUNDARY PERCEPTION

描述：（改编自研究者摘要）

DESCRIPTION: (Adapted from the Investigator's Abstract) The ability to perceive the sharpness of abject boundaries is central to the quality of vision yet very little is know about the underlying neural mechanisms. The broad long-term objective of this research proposal is to reach a more profound understanding of neural mechanisms that determine the perceived form of moving objects in human vision. Under normal viewing conditions, moving objects appear much less blurred than what one would predict from the long duration of visual persistence. This phenomenon is known as motion deblurring. A specific goal of this research is to study the mechanisms underlying motion deblurring and their implications for the perceived form of moving objects. The approach will combine computational and psychophysical methods to test the mechanisms proposed in a neural network model of retino-cortical dynamics. The model leads to the following specific hypotheses: Hypothesis 1 (spatio-temporal profiles): (i) The metacontrast masking function for spatially localized stimuli is oscillatory; (ii) these oscillations are a by-product of the retino-cortical system that occur when it is driven externally by high luminance inputs or internally by focused-arousal/attention; and (iii) the smooth character of the "classical" metacontrast function-extensively reported in the literature-results from spatio-temporal averaging in the post-retinal network. Hypothesis 2 (spatial extent of motion blur): The primary mechanism that determines the length of perceived smear for moving targets is an inhibition from transient cells to sustained cells. Hypothesis 3 (perceived form of motion blur): The brightness profile produced by the retino-cortical dynamics model in response to an isolated moving target will match the psychophysically measured brightness profile of the phenomenon known as "Charpentier's bands."