Simultaneous determination of depth and motion in early vision

摘要

We investigated computationally the role of V1 complex cells in the simultaneous determination of depth and motion, in conjunction with spatiotemporal response characteristics of the cells. First, we examined whether a pooling of simple-cell units with monocular, spatiotemporal response profiles evokes binocular, spatiotemporal response profiles of complex cells. Second, we examined whether a population of the cascade complex-cell models yields the responses being consistent with the perception of depth and motion. We tested various stimuli including ambiguous cases that consist of interocular time delay without spatial disparity. Our simulation study showed that a pooling of simple cells leads to the generation of binocular response profiles of complex cells, and that a population of the complex cells provides a basis for the simultaneous perception of depth and motion.