Frontoparallel Motion Research Articles

Responses in extrastriate cortex to optic flow during simulated turns.

An observer locomoting along a straight path sees a pattern of optic flow in which images move approximately radially outward from his heading point. If the observer turns, the optic flow field changes markedly because each object's image now has a component of horizontal motion added to its "optic flow" motion. We tested the responses of 326 cells in the cat's extrastriate area LS (lateral suprasylvian visual area) to movies simulating the optic flow seen during locomotion in a straight line, and during various simulated turns to the left and right. About 60% of 326 cells tested responded to optic flow simulating turns. Of most interest was a subset of cells, 15% of the total, that had "turn-selective" responses. They responded significantly better to turns in a particular direction (usually to the contralateral side) than to turns in the opposite direction or to optic flow simulating straight-ahead locomotion. For each cell, we generated a display of fronto-parallel motion with a direction and speed that matched the image motion in the preferred turn movie, as seen at the receptive-field center. Most turn-selective cells responded significantly better to their preferred turn movie than to this fronto-parallel stimulus. We examined the role of cells' selectivity for stimulus direction, speed, and acceleration in determining cell preference for particular turns. Direction preference played some role for most cells, but about a third of the cells preferred turn movies that did not reflect their direction selectivity. Other factors, including the presence of opposing motion, only rarely appeared to determine cell preferences for particular turn movies.

Segmentation from motion: combining Gabor- and Mallat-wavelets to overcome the aperture and correspondence problems

A new method for segmentation from motion is presented, which is designed to be part of a general object-recognition system. The key idea is to integrate information from Gabor- and Mallat-wavelet transforms of an image sequence to overcome the aperture and the correspondence problem. It is assumed that objects move fronto-parallel. Gabor-wavelet responses allow accurate estimation of image flow vectors with low spatial resolution. A histogram over this image flow field is evaluated and its local maxima provide a set of motion hypotheses. These serve to reduce the correspondence problem occurring in utilizing the Mallat-wavelet transform, which provides the required high spatial resolution in segmentation. Segmentation reliability is improved by integration over time. The system can segment several small, disconnected, and openworked objects, such as dot patterns. Several examples demonstrate the performance of the system and show that the algorithm behaves reasonably well, even if the assumption of fronto-parallel motion is not met.

Neuronal responsiveness to three-dimensional motion in cat posteromedial lateral suprasylvian cortex.

The neuronal responsiveness to three-dimensional (3D) motion in cat posteromedial lateral suprasylvian (PMLS) cortex was studied using a computer-controlled, stereoscopic 3D graphic display capable of reproducing the major visual cues for natural 3D motion, including motion disparity, size, texture, and shading changes. The animals were anesthetized with nitrous oxide supplemented with alphaxalone, and paralysis prevented eye movement. Systematic investigation of neuronal responsiveness to 3D motions in 26 different directions revealed that more than half of the PMLS cells were selectively responsive to approaching (AP cells, 112 of 271) or recessive motion (RC cells, 64 of 271). The remaining cells were selectively responsive to frontoparallel motion (FP cells, 49 of 271) or nonselectively responsive to motion in multiple directions (NS cells, 46 of 271). The dependency on these visual cues was investigated as a reduction in the response amplitude or the response selectivity for the removal of a single cue from the motion stimuli containing the full visual cues. The AP and RC cells showed a strong dependency on the motion disparity cue, moderate dependency on the size cue, and weak dependency on the texture and shading cues. The FP cells showed no dependency on those visual cues. The cue dependency analysis indicated the existence of nonlinear interactions between those visual cues. Comparison of the responses to a combination of the motion disparity and size cues with the summed responses to each of the individual cues revealed that the responses to the combined cues are roughly predicted as a linear sum between the preferred responses. This comparison also showed nonlinear summation between the nonpreferred responses, i.e., responses to the combined cues were smaller than the summed responses. A similar quasilinear summation of the preferred responses between the two eyes and a nonlinear summation of the nonpreferred responses were found in the AP and RC cells for the motion disparity stimulus. All of these observations indicate that quasilinear and nonlinear interactions of the responses to various stimulus elements underlie the 3D motion responsiveness of the PMLS cells.

Effect of Eye Movements on the Magnitude of fMRI Responses in Extrastriate Cortex during Visual Motion Perception

We have studied the effects of eye movements on the functional magnetic resonance imaging (fMRI) response during global dot motion in extrastriate temporo-occipital cortex. Susceptibility-based functional magnetic resonance, applying gradient-recalled echo-planar imaging (EPI) sequences, was conducted, in 12 volunteers who viewed episodes of moving dots. Anatomical and functional images were recorded with a 1.5 T Siemens Magnetom Vision (TR 5000 ms, TE 70 ms, flip angle 90°, FOV 250 mm, matrix 112 × 128, twelve 4-mm slices parallel to ac — pc plane, ranging from z=-14 – 34 mm). Sagittal T1-weighted images were acquired to determine the ac — pc plane and important anatomical landmarks. Eye movements were monitored with the use of MR-compatible EOG electrodes. Stimuli were back-projected onto a translucent screen and viewed through prisms. The four test conditions were: fronto-parallel motion, during fixation and during target pursuit; expansion-contraction motion, also during fixation and during pursuit. Pixelwise t-tests were performed on functional series recorded during rest and stimulation periods (50 s each) to identify significantly activated voxels. Planes −8 to +8 mm around the ac — pc plane and 20 – 28 mm superior to ac — pc were activated most by visual motion. In addition to primary visual cortical areas, a ventral area in the temporo-occipital cortex (human MT) responded well to both types of visual motion and these responses were enhanced by eye movements. Human MST, located dorsal of MT, showed activation by visual motion during fixation. Its response was also enhanced by pursuit.

Restricted ability to recover three-dimensional global motion from one-dimensional motion signals: Psychophysical observations

We tested human ability to recover the 3D structure and motion information from time-varying images where only 1D motion cues were available. Under these conditions, observers exhibit poor performance in discriminating between two perpendicular axes of rotation, or discriminating between rigid and non-rigid 3D motion. This behavior of the visual system is to be contrasted with the good depth from motion performance exhibited when 2D motion cues are given in the image, as was found previously in numerous studies, and also in the work presented here. In a related paper, we suggest a theoretical framework in which to understand this differential performance on the basis of the two types of motion cues (1D vs 2D). Our findings are consistent with those of previous studies of frontoparallel motion, where it was shown that in many cases, the 1D cues alone were not integrated by the visual system into the correct global motion percept. This accumulating evidence suggests that oriented (1D) motion detectors alone cannot account for observed human performance of global motion perception, and that the role of units such as point or endpoint detectors should be studied further.

The Relationship between Stereoacuity and Stereomotion Thresholds

There are in principle at least two binocular sources of information that could be used to determine the motion of an object towards or away from an observer; such motion produces changes in binocular disparities over time and also generates different image velocities in the two eyes. It has been argued in the past that stereomotion is detected by a mechanism that is independent of that which detects static disparities. More recently this conclusion has been questioned. If stereomotion detection in fact depends upon detecting disparities, there should be a clear correlation between static stereo-detection thresholds and stereomotion thresholds. If the systems are separate, there need be no such correlation. Four types of threshold measurement were performed by means of random-dot stereograms: (1) static stereo detection/discrimination; (2) stereomotion detection in random-dot stereograms (temporally uncorrelated); (3) stereomotion detection in temporally correlated random-dot stereograms; and (4) binocular detection of frontoparallel motion. Three normal subjects and five subjects with unusually high stereoacuities were studied. In addition, two manipulations were performed that altered stereomotion thresholds: changes in mean disparity, and image defocus produced by positive spectacle lenses. Across subjects and conditions, stereomotion thresholds were well correlated with stereo-discrimination thresholds. Stereomotion was poorly correlated with binocular frontoparallel-motion thresholds. These results suggest that stereomotion is detected by means of registering changes in the output of the same disparity detectors that are used to detect static disparities.

Asymmetries in the sensitivity to motion in depth: a centripetal bias.

It is reasonable to ask whether observers are more sensitive to the pattern of image motion caused by forward locomotion through the environment than to the pattern caused by backward locomotion. The distribution of sensitivities of cells in MT does show such a bias, although this bias is minimal at small eccentricities. Additionally, both locomotion-induced stimulation and the sensitivities of MT cells suggest greater sensitivity should be obtained in the lower visual field. Previous research on this issue has usually employed frontoparallel motion in patterns presented to one side of the fixation point. Both centrifugal and centripetal biases have been obtained. In this study the stimuli present motion signals that travel radially from (or towards) the fixation point. These stimuli, which produce a strong percept of motion in depth, are an adaptation of the global-dot-motion stimulus employed by Newsome and Pare. With these stimuli we find that sensitivity to motion in depth is greater in the lower visual field than in the upper visual field, and that sensitivity is greater to centripetal motion than to either centrifugal or frontoparallel motion. This centrifugal bias in sensitivity decreases with eccentricity. The last two findings contradict the notion that the bias is produced by the visual experience induced by normal forward locomotion and also that the detection of motion in depth is subserved by MT.

Functional differentiation between the anterior and posterior Clare-Bishop cortex of the cat

A total of 783 cells were studied extracellularly in anterior (A10-13), posterior (A4-8), and intermediate regions (A8.1-9.9) of Clare-Bishop (CB) cortex of the cat, which were defined according to the anteroposterior coordinate of the stereotaxic axis and probably corresponded to the antero- (AMLS), postero-medial lateral suprasylvian cortex (PMLS), and the border region between the two subareas, respectively. The study was conducted under N2O anesthesia supplemented with continuous infusion of short-lasting anesthetics (Saffan, Glaxo or Etomidate, Janssen), using three types of visual stimulators presenting two- (2D) and three-dimensional (3D) motion stimuli, and visual cues contained in the 3D motion. Neuronal responsiveness was essentially similar between the anterior and posterior CB subdivisions. Both areas contained 1) AP, 2) RC and 3) FP cells, selectively responsive to approaching, recessive and fronto-parallel motion, and 4) NS and 5) U cells, nonselectively responsive and unresponsive to any of these motions. However, a quantitative difference was found: 1) In the posterior CB the FP cell population was the largest, and the frequency reduced in the order of AP, NS, RC and U cells, while the largest population in the anterior CB consisted of the AP and U cells, and the frequency reduced in the order of FP, RC and NS cells. 2) 3D (AP and RC) cells in the posterior CB responded preferentially to approaching motion at a distal range, while those in the anterior CB preferred motion at a proximal range. 3) The 3D cells in the posterior CB were more sensitive to the motion cue and demonstrated lower thresholds for the size cue than the anterior CB cells. 4) The anterior CB cells generally demonstrated high-pass velocity tuning (cut-off around 10 degrees/s) for monoclonal 2D stimulation, while the posterior CB cells demonstrated a broad band-pass tuning (4-120 degrees/s). These findings suggest functional differentiation in neuronal representation of 3D motion signals between the two subdivisions of CB cortex.

The responsiveness of Clare-Bishop neurons to motion cues for motion stereopsis

Under paralysis of eye movement and optical adjustment of visual axes of the two eyes, neuronal responsiveness in cat Clare-Bishop (CB) cortex to the binocular presentation of visual stimuli was studied using two types of visual stimulator which presented: (1) 3-dimensional motion of a visual stimulus; and (2) the motion cues (movement of retinal images in the two eyes) contained in the 3-dimensional motion. On the basis of the responsiveness to 3-dimensional motion presented by the first type of stimulator, the CB cells were classified into approaching (AP) cells which were selectively responsive to the approaching motion of a visual stimulus, recessive (RC) cells responsive to the recessive motion, frontoparallel (FP) cells responsive to the frontoparallel motion and non-selective (NS) cells responsive to more than two types of motion. The investigation of the CB cells with 2-dimensional motion stimulus demonstrated 3 different types of binocular interaction: facilitatory (52/239); antagonistic (33/239) or linear summation (154/239). Cells exhibiting the facilitatory interaction (n=52) were all FP cells, those exhibiting the antagonistic interaction were either AP (n=25) or RC cells (n=8), and those exhibiting the linear summation were comprised of all varieties of cells (49 AP, 17 RC, 31 FP and 57 NS cells). The cells responsive to the approaching (37 AP cells) or recessive motion between the center of the receptive area and the nose (10 RC cells), or frontoparallel motion (42 FP cells) in the horizontal direction exhibited the selective responsiveness to the motion disparity (a combination of horizontal movement of retinal images in the two eyes), and those responsive to the vertically or obliquely deviating approaching (20 AP cells) or recessive motion (14 RC cells), or frontoparallel motion in the vertical or oblique directions (13 FP cells) exhibited that for a combination of the motion disparity and the frontoparallel motion in the vertical or oblique direction. These findings indicate that the CB cell responsiveness to 3-dimensional motion of a visual stimulus is explained by the binocular intergration of motion signal viewed by each eye.

The responsiveness of Clare-Bishop neurons to motion cues for motion stereopsis

Under paralysis of eye movement and optical adjustment of visual axes of the two eyes, neuronal responsiveness in cat Clare-Bishop (CB) cortex to the binocular presentation of visual stimuli was studied using two types of visual stimulator which presented: (1) 3-dimensional motion of a visual stimulus; and (2) the motion cues (movement of retinal images in the two eyes) contained in the 3-dimensional motion. On the basis of the responsiveness to 3-dimensional motion presented by the first type of stimulator, the CB cells were classified into approaching (AP) cells which were selectively responsive to the approaching motion of a visual stimulus, recessive (RC) cells responsive to the recessive motion, frontoparallel (FP) cells responsive to the frontoparallel motion and non-selective (NS) cells responsive to more than two types of motion. The investigation of the CB cells with 2-dimensional motion stimulus demonstrated 3 different types of binocular interaction: facilitatory ( 52 239 ); antagonistic ( 33 239 ) or linear summation ( 154 239 ). Cells exhibiting the facilitatory interaction ( n = 52) were all FP cells, those exhibiting the antagonistic interaction were either AP ( n = 25) or RC cells ( n = 8), and those exhibiting the linear summation were comprised of all varieties of cells (49 AP, 17 RC, 31 FP and 57 NS cells). The cells responsive to the approaching (37 AP cells) or recessive motion between the center of the receptive area and the nose (10 RC cells), or frontoparallel motion (42 FP cells) in the horizontal direction exhibited the selective responsiveness to the motion disparity (a combination of horizontal movement of retinal images in the two eyes), and those responsive to the vertically or obliquely deviating approaching (20 AP cells) or recessive motion (14 RC cells), or frontoparallel motion in the vertical or oblique directions (13 FP cells) exhibited that for a combination of the motion disparity and the frontoparallel motion in the vertical or oblique direction. These findings indicate that the CB cell responsiveness to 3-dimensional motion of a visual stimulus is explained by the binocular integration of motion signal viewed by each eye.

The responsiveness of Clare-Bishop neurons to size cues for motion stereopsis

The responsiveness to the size cue, and a combination of the size and motion cues contained in 3-dimensional motion of a visual stimulus was studied in 118 Clare-Bishop (CB) cells, including 37 and 10 cells selectively responsive to approaching (AP) and to recessive motion along the the axis through the center of the receptive area and the nose (RC), 40 cells responsive to the frontoparallel motion in the horizontal direction (FP), 23 cells rather non-selectively responsive to the two types of motion (NS) and 8 cells responsive to the size cue but unresponsive to the 3-dimensional motion (SZ). About three quarters of the AP cells (27/37) were responsive to both an increase in the stimulus size and divergent motion of the retinal images in the two eyes, which represents the size and motion cues for the approaching motion along the axis through the center of the receptive area and the nose, and were optimally excited by a combination of the two visual cues. About the same fraction of the RC cells (6/10) was responsive to the motion (convergent motion) and size cues (a decrease in stimulus size), and optimally excited by a combination of the two visual cues. In contrast, only a small fraction of the FP cells (6/40) were sensitive to the size cue, and all FP cells were optimally excited by the single presentation of the motion cue for the frontoparallel motion (either right- or leftward motion in both eyes). Responses were frequently (18/40) smaller for the combined presentation of the two visual cues than for the single presentation of the motion cue. Similarly, a small fraction of the NS cells (7/23) was sensitive to the size cue, but many of them (16/23) were non-selectively responsive to the size cues for the approaching and recessive motion. A similar study in an additional 108 (78 AP and 30 RC) cells which were responsive to approaching or recessive motion vertically or obliquely deviating from the axis through the center of the receptive area and the nose demonstrated that they were also sensitive to both motion and size cues for that approaching or recessive motion, and were optimally excited by a combination of the two visual cues. These findings indicate that the CB cell responsiveness to 3-dimensional motion is based on the integration of the motion and size signals conveyed through the two eyes. A model of neuronal circuitry was constructed to explain the CB cell responsiveness to the motion and size signals.

The responsiveness of Clare-Bishop neurons to size cues for motion stereopsis

The responsiveness to the size cue, and a combination of the size and motion cues contained in 3-dimensional motion of a visual stimulus was studied in 118 Clare-Bishop (CB) cells, including 37 and 10 cells selectively responsive to approaching (AP) and to recessive motion along the the axis through the center of the receptive area and the nose (RC), 40 cells responsive to the frontoparallel motion in the horizontal direction (FP), 23 cells rather non-selectively responsive to the two types of motion (NS) and 8 cells responsive to the size cue but unresponsive to the 3-dimensional motion (SZ). About three quarters of the AP cells ( 27 37 ) were responsive to both an increase in the stimulus size and divergent motion of the retinal images in the two eyes, which represents the size and motion cues for the approaching motion along the axis through the center of the receptive area and the nose, and were optimally excited by a combination of the two visual cues. About the same fraction of the RC cells ( 6 10 ) was responsive to the motion (convergent motion) and size cues (a decrease in stimulus size), and optimally excited by a combination of the two visual cues. In contrast, only a small fraction of the FP cells ( 6 40 ) were sensitive to the size cue, and all FP cells were optimally excited by the single presentation of the motion cue for the frontoparallel motion (either right- or leftward motion in both eyes). Responses were frequently ( 18 40 ) smaller for the combined presentation of the two visual cues than for the single presentation of the motion cue. Similarly, a small fraction of the NS cells ( 7 23 ) was sensitive to the size cue, but many of them ( 16 23 ) were non-selectively responsive to the size cues for the approaching and recessive motion. A similar study in an additional 108 (78 AP and 30 RC) cells which were responsive to approaching or recessive motion vertically or obliquely deviating from the axis through the center of the receptive area and the nose demonstrated that they were also sensitive to both motion and size cues for that approaching or recessive motion, and were optimally excited by a combination of the two visual cues. These findings indicate that the CB cell responsiveness to 3-dimensional motion is based on the integration of the motion and size signals conveyed through the two eyes. A model of neuronal circuitry was constructed to explain the CB cell responsiveness to the motion and size signals.

Responsiveness of Clare-Bishop neurons to visual cues associated with motion of a visual stimulus in three-dimensional space

Photic responsiveness of cells in the medial bank of the lateral suprasylvian cortex (Clare-Bishop area) was studied using a three-dimensional visual stimulator that reproduced two visual cues (motion disparity and change in size) for perception of three-dimensional motion of a visual stimulus. About one third of them (48/148) were selectively responsive to motion disparity corresponding to approaching (AP cells, n = 30) or recessive motion (RC cells, N = 18), another half to motion of retinal images in the same direction between the two eyes corresponding to fronto-parallel motion (FP cells, n = 75), and the remaining cells were rather equally responsive to these types of stimuli (NS cells, n = 25). More than a half of the AP (19/30) or RC (11/18) cells were also responsive to increase or decrease in stimulus size, respectively, and they were optimally activated by a combination of the motion and size stimuli while relatively few FP and NS cells were sensitive to change in stimulus size. These findings indicate that the Clare-Bishop cells encode three-dimensional motion on the basis of photic responsiveness to the motion and size cues.

Functional properties of neurons in middle temporal visual area of the macaque monkey. II. Binocular interactions and sensitivity to binocular disparity.

1. Electrophysiological recordings were made in the middle temporal visual area (MT) of five macaque monkeys. Binocularity and selectivity for disparity were examined using a computer-driven stimulator to activate each eye independently. Results were obtained from 91 single units in MT. 2. Most units in MT receive approximately balanced inputs from the two eyes, and very few could be driven through only one eye. 3. In one type of test for disparity selectivity, units were examined with stimuli that had different but fixed horizontal disparities, thereby simulating frontoparallel movements at different distances from the animal. About two-thirds of ;the units tested for fixed disparity selectivity (52/76) showed pronounced sensitivity to horizontal disparity. Most of these units could be grouped into the same four classes of disparity-tuned units that have previously been described in V1 and V2 of the macaque: near, far, tuned excitatory, and tuned inhibitory. 4. Twenty units were tested for sensitivity to vertical stimulus disparity, which does not normally contribute to stereopsis. Most were as sensitive to vertical disparities as to horizontal. 5. Units were also tested for selectivity for stimuli that moved with changing disparity, simulating trajectories with components of motion toward or away from the animal (motion in depth). No units were found to be truly selective for motion in depth. Units tuned for fixed disparity could appear to prefer motion in depth if tested only with trajectories whose common center point was far from the unit's optimal fixed disparity. However, we do not consider this to represent genuine selectivity for motion in depth, since 1) the responses ara adequately and more easily explained in terms of selectivity for fixed disparity and 2) the best overall response of these units is to frontoparallel motion at the optimal fixed disparity. This observation bears importantly on the interpretation of motion in depth selectivity in previous investigations. 6. The presence of a substantial degree of selectivity for fixed disparity in MT, together with previously demonstrated selectivities for direction and speed, indicates that MT is well suited for the analysis of motion in three-dimensional space.