The Informational Basis of Motion Coherence

Battle of the Mondrians: Investigating the Role of Unpredictability in Continuous Flash Suppression

In continuous flash suppression (CFS), a dynamic sequence of Mondrian patterns presented to one eye suppresses a static target in the other eye for several seconds at a time. Its effectiveness has been linked to low-level properties such as spatial frequency and orientation, but the role of higher order influences remains unstudied. Here, using a tracking paradigm, we asked if the spatial and temporal predictability of the Mondrian sequence affects CFS dynamics. Predictable temporal sequences were regularly updated every 100 ms or modulated sinusoidally in pixel luminance at 2 Hz. Unpredictable temporal sequences were irregularly updated or had stochastic pixel luminance changes across time. To vary spatial predictability, sequences were either updated with different Mondrian patterns or had a fixed spatial pattern. We found a modest effect of spatial uncertainty when the target modulation was low (0.125 Hz) but not temporal uncertainty, which had no significant effects regardless of target modulation. Similar results were obtained when we pitted the standard Mondrian sequence against sequences with a fixed spatial pattern and temporally low-pass filtered sequences in a binocular rivalry paradigm. Thus, not only was the effect of information predictability was modest and spatial, but it was also dependent on the presence of higher temporal frequencies. Together, the results demonstrate the significance of low-level properties in affecting CFS dynamics and the possible involvement of pattern structure masking in CFS.

Pooling elementary motion signals into perception of global motion direction

Six observers were asked to indicate in which of two opposite directions, to the right or to the left, an entire display appeared to move, based on the proportion of right vs leftward motion elements, each of which was distinctly visible. The performance of each observer was described by Thurstone's discriminative processes and Bernoulli trial models which described empirical psychometric functions equally well. Although formally it was impossible to discriminate between these two models, treating observer as a counting device which measures a randomly selected subsample of all available motion elements had certain advantages. According to the Bernoulli trial model decisions about the global motion direction in a range of 12-800 elements were based on taking into account about 4±2 random moving dot elements. This small number is not due to cancellation of the opposite motion vectors since the motion direction recognition performance did not improve after the compared motion directions were made orthogonal. This may indicate that the motion pooling mechanism studied in our experiment is strongly limited in capacity.

Attentional limitations in the sensing of motion direction

Attentional constraints in the perceptual analysis of motion direction were examined using two independent paradigms: redundant target visual search and the analysis of fluctuations in discrimination accuracy at threshold. Results from both methods implied that directions of object motion are analyzed in parallel when those motions are translations, independent of the observer's line of sight. The registration of rotation direction appears to be subject to a qualitatively different protocol, one that is highly capacity limited and serial-like. These results suggest that scene-based descriptions, as opposed to image-based descriptions of motion, mediate the allocation of attention.

Visual form created solely from temporal structure

In several experiments, it was found that global perception of spatial form can arise exclusively from unpredictable but synchronized changes among local features. Within an array of nonoverlapping apertures, contours move in one of two directions, with direction reversing randomly over time. When contours within a region of the array reverse directions in synchrony, they stand out conspicuously from the rest of the array where direction reversals are unsynchronized. Clarity of spatial structure from synchronized change depends on the rate of motion reversal and on the proportion of elements reversing direction in synchrony. Evidently, human vision is sensitive to the rich temporal structure in these stochastic events.

Perception of biological motion

Boundary conditions for perception of biological motion were explored with the use of computer-generated point-light animation sequences. Perception of this unique form of structure from motion is immune to variations in dot contrast polarity, dot disparity, and spatial-frequency filtering. Biological motion is perceived in texture-defined animation sequences that presumably stimulate only second-order motion pathways, and it is undisturbed by dichoptic presentation of portions of the animation tokens separately to the two eyes.