The human visual system is foveated: we can see fine spatial details in central vision, whereas resolution is poor in our peripheral visual field, and this loss of resolution follows an approximately logarithmic decrease. Additionally, our brain organizes visual input in polar coordinates. Therefore, the image projection occurring between retina and primary visual cortex can be mathematically described by the log-polar transform. Here, we test and model how this space-variant visual processing affects how we process binocular disparity, a key component of human depth perception. We observe that the fovea preferentially processes disparities at fine spatial scales, whereas the visual periphery is tuned for coarse spatial scales, in line with the naturally occurring distributions of depths and disparities in the real-world. We further show that the visual system integrates disparity information across the visual field, in a near-optimal fashion. We develop a foveated, log-polar model that mimics the processing of depth information in primary visual cortex and that can process disparity directly in the cortical domain representation. This model takes real images as input and recreates the observed topography of human disparity sensitivity. Our findings support the notion that our foveated, binocular visual system has been moulded by the statistics of our visual environment.

We have developed a low-cost, practical gaze-contingent display in which natural images are presented to the observer with dioptric blur and stereoscopic disparity that are dependent on the three-dimensional structure of natural scenes. Our system simulates a distribution of retinal blur and depth similar to that experienced in real-world viewing conditions by emmetropic observers. We implemented the system using light-field photographs taken with a plenoptic camera which supports digital refocusing anywhere in the images. We coupled this capability with an eye-tracking system and stereoscopic rendering. With this display, we examine how the time course of binocular fusion depends on depth cues from blur and stereoscopic disparity in naturalistic images. Our results show that disparity and peripheral blur interact to modify eye-movement behavior and facilitate binocular fusion, and the greatest benefit was gained by observers who struggled most to achieve fusion. Even though plenoptic images do not replicate an individual’s aberrations, the results demonstrate that a naturalistic distribution of depth-dependent blur may improve 3-D virtual reality, and that interruptions of this pattern (e.g., with intraocular lenses) which flatten the distribution of retinal blur may adversely affect binocular fusion.

A computational model for the control of horizontal vergence, based on a population of disparity tuned complex cells, is presented. Since the population is able to extract the disparity map only in a limited range, using the map to drive vergence control means to limit its functionality inside this range. The model directly extracts the disparity-vergence response by combining the outputs of the disparity detectors without explicit calculation of the disparity map. The resulting vergence control yields to stable fixation and has small response time to a wide range of disparities. Experimental simulations with synthetic stimuli in depth validated the approach

Motion estimation has been studied extensively in neuroscience in the last two decades. Even though there has been some early interaction between the biological and computer vision communities at a modelling level, comparatively little work has been done on the examination or extension of the biological models in terms of their engineering efficacy on modern optical flow estimation datasets. An essential contribution of this paper is to show how a neural model can be enriched to deal with real sequences. We start from a classical V1-MT feedforward architecture. We model V1 cells by motion energy (based on spatio-temporal filtering), and MT pattern cells (by pooling V1 cell responses). The efficacy of this architecture and its inherent limitations in the case of real videos are not known. To answer this question, we propose a velocity space sampling of MT neurons (using a decoding scheme to obtain the local velocity from their activity) coupled with a multi-scale approach. After this, we explore the performance of our model on the Middlebury dataset. To the best of our knowledge, this is the only neural model in this dataset. The results are promising and suggest several possible improvements, in particular to better deal with discontinuities. Overall, this work provides a baseline for future developments of bio-inspired scalable computer vision algorithms and the code is publicly available to encourage research in this direction

No abstract