Back
Contrast-reversed binocular dot-pairs in random-dot stereograms for depth per-ception in central visual field: Probing the dynamics of feedforward-feedback pro-cesses in visual inference
{In a random-dot stereogram (RDS), the percept of object surfaces in a three-dimensional scene is generatedby images presented to left and right eyes that comprise interocularly corresponding random black and white dots. The spatialdisparities between the binocularly corresponding dots determine the depths of object surfaces. If the dots are contrast-reversed,such that a black dot in one monocular image corresponds to a white dot in the other, disparity-tuned neurons in primary visualcortex (V1) respond as if their preferred disparities become non-preferred and vice versa, thereby reversing the disparity signreported to higher visual areas. Reversed depth is imperceptible in the central visual field (although perceptible peripherally).This study demonstrates that, in central vision, adding reversed depth signals to a noisy RDS can augment or degrade depthperception. Augmentation occurs when the reversed depth signals (from contrast-reversed dots) are congruent, or agree, withthe normal depth signals (from the contrast-matched dots) on the reported disparity sign. Degradation occurs when the reversedand the normal signals are incongruent with each otherandwhen the RDS images are sufficiently brief. These findings areconsistent with the recently proposed central-peripheraldichotomy, which suggests that central vision (relative toperipheralvision) has a stronger top-down feedback from higher to lower brain areas to disambiguate noisy and ambiguous inputs from V1.Through an analysis-by-synthesis computation, this feedback vetoes misleading V1 signals, or completes and repairs imperfectV1 signals. When the RDS is viewed too briefly to allow time forfeedback, both augmentation and degradation work by addingthe reversed signals to the normal signals in the feedforward direction. With a sufficient viewing duration, the feedback removesthe degradation but preserves the augmentation as follows.When the reversed depth signals are incongruent, the contrast-reverseddots are no worse than monocular noise dots in influencing perception, and in particular the reversed depth signal is imperceptiblewhen normal depth signals are absent. When the reversed depth signals are congruent, they augment the percept of the normaldepth signals as if the deficiencies in the reversed signals are filled in or corrected by the analysis-by-synthesis in thefeedbackprocess. The perceptual impact of the reversed depth signals reflects the underlying dynamics in the Feedforward-Feedback-Verify-and-reWeight (FFVW) process for visual inference in central vision.}
@article{item_3291577, title = {{Contrast-reversed binocular dot-pairs in random-dot stereograms for depth per-ception in central visual field: Probing the dynamics of feedforward-feedback pro-cesses in visual inference}}, journal = {{Vision Research}}, abstract = {{In a random-dot stereogram (RDS), the percept of object surfaces in a three-dimensional scene is generatedby images presented to left and right eyes that comprise interocularly corresponding random black and white dots. The spatialdisparities between the binocularly corresponding dots determine the depths of object surfaces. If the dots are contrast-reversed,such that a black dot in one monocular image corresponds to a white dot in the other, disparity-tuned neurons in primary visualcortex (V1) respond as if their preferred disparities become non-preferred and vice versa, thereby reversing the disparity signreported to higher visual areas. Reversed depth is imperceptible in the central visual field (although perceptible peripherally).This study demonstrates that, in central vision, adding reversed depth signals to a noisy RDS can augment or degrade depthperception. Augmentation occurs when the reversed depth signals (from contrast-reversed dots) are congruent, or agree, withthe normal depth signals (from the contrast-matched dots) on the reported disparity sign. Degradation occurs when the reversedand the normal signals are incongruent with each otherandwhen the RDS images are sufficiently brief. These findings areconsistent with the recently proposed central-peripheraldichotomy, which suggests that central vision (relative toperipheralvision) has a stronger top-down feedback from higher to lower brain areas to disambiguate noisy and ambiguous inputs from V1.Through an analysis-by-synthesis computation, this feedback vetoes misleading V1 signals, or completes and repairs imperfectV1 signals. When the RDS is viewed too briefly to allow time forfeedback, both augmentation and degradation work by addingthe reversed signals to the normal signals in the feedforward direction. With a sufficient viewing duration, the feedback removesthe degradation but preserves the augmentation as follows.When the reversed depth signals are incongruent, the contrast-reverseddots are no worse than monocular noise dots in influencing perception, and in particular the reversed depth signal is imperceptiblewhen normal depth signals are absent. When the reversed depth signals are congruent, they augment the percept of the normaldepth signals as if the deficiencies in the reversed signals are filled in or corrected by the analysis-by-synthesis in thefeedbackprocess. The perceptual impact of the reversed depth signals reflects the underlying dynamics in the Feedforward-Feedback-Verify-and-reWeight (FFVW) process for visual inference in central vision.}}, publisher = {Pergamon}, address = {Amsterdam}, year = {2021}, slug = {item_3291577}, author = {Zhaoping, L} }