Perceptual depth synthesis in the visual system as revealed by selective adaptation

Selective adaptations was used to determine the degree of interactions between channels processing relative depth from stereopsis, motion parallax, and texture. Monocular adaptations with motion parallax or binocular stationary adaptation caused test surfaces, viewed either stationary binocularly or monocularly with motion parallax, to appear to slant in the opposite direction compared with the slant initially adapted to. Monocular adaptations on frontoparallel surfaces covered with a pattern of texture gradients caused a subsequently viewed test surface, viewed either monocularly with motion parallax or stationary binocularly, to appear to slant in the opposite direction as the slant indicated by the texture in the adaptation condition. No aftereffect emerged in the monocular stationary test condition. A mechanism of independent channels for relative depth perception is dismissed in favor of a view of an asymmetrical interactive processing of different information sources. The results suggest asymmetrical inhibitory interactions among habituating slant detector units receiving inputs from static disparity, dynamic disparity, and texture gradients.     

Texture size specificity in the slant aftereffect

Transfer of the median plane slant aftereffect was assessed across changes in stimulus texture size (sine-wave grating frequency). Under binocular viewing, reliable decrements in aftereffect magnitude were observed when texture size was changed, compared with no-change control conditions. Under monocular viewing conditions, no significant aftereffects were found. The results indicate a spatial-frequency-specific component of binocular slant aftereffects.     

Aftereffect of inspection of a perspectival stimulus for slant depth: A new normalization effect

The Ss made judgments of the frontoparallel plane monocularly and binocularly by adjusting the position of a rotatable thin luminous line viewed in darkness. The judgments were made before and after a 10-min period of monocular inspection of a luminous outline frontoparallel trapezoid which was the projection of a rectangle rotated in depth. Verbal reports indicated normalization of perspectival slant depth, i.e., the inspection figure appeared less rotated, and the line settings exhibited an aftereffect bf normalization. These findings may contribute toward an account of certain previously unexplained observations reported in an earlier study by Epstein and Morgan-Paap (1974) dealing with adaptation to uniocular image magnification.     

The perception of depth and slant from texture in three-dimensional scenes.

The perception of depth and slant in three-dimensional scenes specified by texture was investigated in five experiments. Subjects were presented with computer-generated scenes of a ground and ceiling plane receding in depth. Compression, convergence, and grid textures were examined. The effect of the presence or absence of a gap in the center of the display was also assessed. Under some conditions perceived slant and depth from compression were greater than those found with convergence. The relative effectiveness of compression in specifying surface slant was greater for surfaces closer to ground planes (80 degrees slant) than for surfaces closer to frontal parallel planes (40 degrees slant). The usefulness of compression was also observed with single-plane displays and with displays with surfaces oriented to reduce information regarding the horizon.     

Stereospatial masking and aftereffect with normal and transformed random-dot patterns

Masking and aftereffect in the perception of binocular depth were studied using random-dot sterograms as adaptation and target stimuli. Detection of the target was impaired by prior adaptation only when the two stimuli differed in disparity by less than 2 minarc. The masking function was unaffected by uniocular enlargement and blurring within the adaptation stimulus, but masking was no longer selective to disparity when the elements seen by the two eyes were reversed in brightness. The stereoscopic depth aftereffect was also insensitive to uniocular enlargement and blurring, and could not be generated when there was brightness complementation within the adaptation stimulus. Both the masking and aftereffect data are interpreted as evidence that stereospatial detectors in human vision are insensitive to transforms that maintain luminance-spatial correlations in binocular input.     

Slant perception and binocular brightness differences: Some aftereffects of viewing apparent and objective surface slants

Placing a neutral-density filter in front of one eye produces two kinds of distortion in the perceived slant of a binocularly viewed rotating disk: (1) the top or the bottom of a disk rotating in a frontoparallel plane appears displaced toward or away from the observer, depending on the direction of rotation and whether the left or right eye is filtered; and (2) the left or right side of such disk—rotating or stationary—appears closer, depending on whether the left or right eye is filtered. The Pulfrich phenomenon accounts for the first variety of apparent slant, and the Venetian blind effect accounts for the second. Viewing the apparent slant of the rotating disk produces an aftereffect of slant in the third dimension which is greater than the aftereffect of viewing an objective slant of the same direction and magnitude.     

Aftereffect of high-speed motion.

A visual illusion known as the motion aftereffect is considered to be the perceptual manifestation of motion sensors that are recovering from adaptation. This aftereffect can be obtained for a specific range of adaptation speeds with its magnitude generally peaking for speeds around 3 deg s-1. The classic motion aftereffect is usually measured with a static test pattern. Here, we measured the magnitude of the motion aftereffect for a large range of velocities covering also higher speeds, using both static and dynamic test patterns. The results suggest that at least two (sub)populations of motion-sensitive neurons underlie these motion aftereffects. One population shows itself under static test conditions and is dominant for low adaptation speeds, and the other is prevalent under dynamic test conditions after adaptation to high speeds. The dynamic motion aftereffect can be perceived for adaptation speeds up to three times as fast as the static motion aftereffect. We tested predictions that follow from the hypothesised division in neuronal substrates. We found that for exactly the same adaptation conditions (oppositely directed transparent motion with different speeds), the aftereffect direction differs by 180 degrees depending on the test pattern. The motion aftereffect is opposite to the pattern moving at low speed when the test pattern is static, and opposite to the high-speed pattern for a dynamic test pattern. The determining factor is the combination of adaptation speed and type of test pattern.     

Adaptation to telestereoscopic viewing measured by one-handed ball-catching performance

A one-handed ball-catching task was used to study the disturbance of depth judgement induced by telestereoscopic viewing (ie viewing with increased effective interocular separation), the recovery of performance with experience in the telestereoscope, and the errors that subsequently arose when the telestereoscope was removed. The ball's trajectory was variable so that subjects had to control both the position and the timing of the grasp in order to catch the ball. On first wearing the telestereoscope, subjects closed the hand when the ball was approximately twice as far away from the eyes as the hand was. After fewer than twenty trials in the telestereoscope subjects were closing the hand at approximately the correct time and place, although rather more trials were needed for ball-catching performance to recover to normal. When the telestereoscope was removed there was an aftereffect, with subjects making the opposite errors to when they began the task. The existence of an aftereffect shows that the process of adaptation involves reevaluation rather than neglect of the misleading binocular information. Helmholtz's theory that telestereoscopes cause the world to be perceived as a scale model is considered. Initial misreaching is roughly consistent with this theory, but there are insufficient data to test it rigorously. Data from the aftereffect phase are clearly inconsistent with the theory. The results confirm the importance of binocular information in dynamic motor tasks, such as ball catching.     

A three-dimensional motion aftereffect produced by prolonged adaptation to a rotation simulation

After adaptation to a perspective simulation of a square plane rotating in depth, an ambiguous rotation simulation (ie one containing no perspective information) appears to rotate in the direction opposite that of adaptation. The strength of this three-dimensional motion aftereffect (MAE) is proportional to the amount of perspective available in the adaptation display and, in the dark, decays to about 75% of its initial strength within about 546 s. The nature of the testing situation and a control experiment suggest that the three-dimensional MAE is not caused by retinal adaptation of two-dimensional directionally selective mechanisms.     

THE GENERALITY OF NEGATIVE AFTEREFFECT FOLLOWING ADAPTATION TO CURVATURE

Twenty-six subjects made judgments of curvature by the method of magnitude estimation. The influence of prior adaptation to a moderate degree of curvature was assessed for test-curvatures greater and less than the adapting curvature and in the same and opposite directions. The median aftereffect was negative in all cases but not statistically significant for the most extreme difference between adapting and test curvatures. It is argued that negative aftereffect for all test-curvatures along the same stimulus-continuum is a necessary condition for postulating that these experimental effects represent a general normalizing tendency in perception.     

Modifying an underlying component of perceived arm length: adaptation of tactile location induced by spatial discordance

In four experiments we examined the adaptation and aftereffect that resulted from a treatment yielding tactile/kinesthetic length discordance between the arms. Perceived discordance diminished with trials and tended to zero. Subsequent visual/tactile cross-modal judgments of distance showed the aftereffect to be a change in the perceived location of an unseen probed spot on each hand with respect to the location of a truly coincident visual marker. This occurred toward the body for the probed spot on one arm and away from the body on the other. There were three other main findings: (a) Arm movement was not a necessary condition for adaptation or aftereffect; (b) with intrinsic length information about the right arm present, but touch information from the right index finger absent during treatment, adaptation and aftereffect were abolished; (c) aftereffects of tactile location that were manifest at the hand and wrist tended to zero when a point close to the elbow was tested with a cross-modal procedure. The experiments provide evidence that the mapping of the tactile sheet onto an internal length domain had been modified by the treatment. The sensory consequences of the treatment led many subjects to report spontaneously that their arms felt to be of different lengths.     

Learning motor synergies makes use of information on muscular load

Prism adaptation, a form of procedural learning, requires the integration of visual and motor information for its proper acquisition. Although the role of the visual feedback has begun to be understood, the nature of the motor information necessary for the development of the adaptation remains unknown. In this work we have tested the idea that modifying the arm load at different stages of the adaptation process, and the ensuing change of motor information perceived by the subjects, would modify the final properties of the adaptation. We trained a set of subjects to throw balls to a target while wearing prism glasses and varied the weight of their arms at different time points during the task. We observed that the acquisition of the adaptation was not affected by the change in load. However, its persistence (i.e., the aftereffect) was reduced when tested under a weight condition different from the training trials. Furthermore, when the training weight conditions were restored later during testing, a second, late aftereffect was unmasked, suggesting that the missing aftereffect did not disappear but had remained latent. Our results show that the internal representation of a motor memory incorporates information about load conditions and that the memory stored under a specific weight condition can be fully retrieved only when the original training condition is restored.     

Adaptation to prism-displaced vision: The importance of target-pointing

Two experiments investigated the hypothesis that the experience of manually pointing at visual targets enhances motoric adaptation to prism-displaced vision. Experiment 1 indicated that when adaptation was measured by means of redirected pointing behavior (negative aftereffect) it varied directly with the specificity of the target, the least adaptation occurring when no target was available. This relationship was not observed when adaptation was measured in terms of a shift in the felt position of the prism-exposed hand (proprioceptive shift). Experiment 2 demonstrated that after double the prism-exposure trials used in Experiment 1, target-pointing experience continued to enhance adaptation (as indexed by both types of adaptation measure). In both experiments negative aftereffect was significantly larger than proprioceptive shift for all experimental conditions and the two measures were not correlated. These latter two findings cast doubt on Harris’s notion that negative aftereffect is entirely the result of altered position sense.     

The interplay between stereopsis and structure from motion

In a series of psychophysical experiments, an adaptation paradigm was employed to study the influence of stereopsis on perception of rotation in an ambiguous kinetic depth (KD) display. Without prior adaptation or stereopsis, a rotating globe undergoes spontaneous reversals in perceived direction of rotation, with successive durations of perceived rotation being random variables. Following 90 sec of viewing a stereoscopic globe undergoing unambiguous rotation, the KD globe appeared to rotate in a direction opposite that experienced during the stereoscopic adaptation period. This adaptation aftereffect was short-lived, and it occurred only when the adaptation and test figures stimulated the same retinal areas, and only when the adaptation and test figures rotated about the same axis. The aftereffect was just as strong when the test and adaptation figures had different shapes, as long as the adaptation figure contained multiple directions of motion imaged at different retinal disparities. Nonstereoscopic adaptation figures had no effect on the perceived direction of rotation of the ambiguous KD figure. These results imply that stereopsis and motion strongly interact in the specification of structure from motion, a result that complements earlier work on this problem.     

The site of binocular rivalry suppression.

Two experiments were performed to localize the site of binocular rivalry suppression in relation to the locus of grating adaptation. In one experiment it was found that phenomenal suppression of a high-contrast adaptation grating presented to one eye had no influence on the strength of the threshold-elevation aftereffect measured interocularly. Evidently information about the adaptation grating arrives at the site of the aftereffect (presumably binocular neurons) even during suppression. In a second experiment 60 s of grating adaptation was found to produce a short-term reduction in the predominance of the adapted eye during binocular rivalry. These findings provide converging lines of evidence that suppression occurs at a site in the human visual system after the locus of grating adaptation and, hence, after the striate cortex.     

视时距知觉适应后效的空间选择性

时距知觉适应后效是指长时间适应于某一特定时距会导致个体对后续时距产生知觉偏差。其中对视时距知觉适应后效空间选择性的探讨存在争议,有研究支持位置不变性,也有研究支持位置特异性。这类研究能有效揭示时距编码的认知神经机制,位置不变性可能意味着时距编码位于较高级的脑区,而位置特异性则可能意味着时距编码位于初级视觉皮层。未来还可以探究时距知觉适应后效的视觉坐标表征方式,开展多通道研究以及相应的神经基础研究。     

Induction of the McCollough effect II: Two different mechanisms

An orientation-specific chromatic aftereffect was observed when a single colored grating was used as an induction stimulus. The magnitude of the aftereffect was compared to that obtained when alternating orthogonal gratings in complementary hues were used as induction stimuli. The two-stimulus condition produced a stronger aftereffect than a single-stimulus condition. This facilitation was also obtained when a colored plain square with no grating was substituted for the second colored grating in the two-stimulus condition. The results suggest that the McCollough effect involves adaptation of two different mechanisms, one which is orientation-specific and one which is not.     

When texture takes precedence over motion in depth perception

Both texture and motion can be strong cues to depth, and estimating slant from texture cues can be considered analogous to calculating slant from motion parallax (Malik and Rosenholtz 1994, report UCB/CSD 93/775, University of California, Berkeley, CA). A series of experiments was conducted to determine the relative weight of texture and motion cues in the perception of planar-surface slant when both texture and motion convey similar information. Stimuli were monocularly viewed images of planar surfaces slanted in depth, defined by texture and motion information that could be varied independently. Slant discrimination biases and thresholds were measured by a method of single-stimuli binary-choice procedure. When the motion and texture cues depicted surfaces of identical slants, it was found that the depth-from-motion information neither reduced slant discrimination thresholds, nor altered slant discrimination bias, compared to texture cues presented alone. When there was a difference in the slant depicted by motion and by texture, perceived slant was determined almost entirely by the texture cue. The regularity of the texture pattern did not affect this weighting. Results are discussed in terms of models of cue combination and previous results with different types of texture and motion information.     

Surface separation decreases stereoscopic slant but a monocular aperture increases it.

When an isolated surface is stereoscopically slanted around its vertical axis, perceived slant is attenuated relative to prediction, whereas when a frontal-plane surface is placed above or below the slanted surface, slant is close to the predicted magnitude. Gillam et al (1988 Journal of Experimental Psychology: Human Perception and Performance 14 163-175) have argued that this slant enhancement is due to the introduction of a gradient of relative disparities across the abutment of the two surfaces which is a more effective stimulus for slant than is the gradient of absolute disparities present when the slanted surface is presented alone. To test this claim we varied the separation between the two surfaces, along either the vertical or depth axis. Since these manipulations have been reported to reduce the depth response to individual relative disparities, they should similarly affect any slant response based on a gradient of relative disparities. As predicted, increasing the separation, vertically or in depth, systematically reduced both the perceived slant of the stereoscopically slanted surface and also the stereo contrast slant induced in the frontal-plane surface. These results are not predicted by alternative accounts of slant enhancement (disparity-gradient contrast, normalisation, frame of reference). We also demonstrated that sidebands of monocular texture, when added to equate the half-image widths of the slanted surface, increased the perceived slant of this surface (particularly when presented alone) and reduced the contrast slant. Monocular texture, by signalling occlusion, appeared to provide absolute slant information which determined how the total relative slant perceived between the surfaces was allocated to each.     

Auditory motion aftereffects

Observers were adapted to simulated auditory movement produced by dynamically varying the interaural time and intensity differences of tones (500 or 2,000 Hz) presented through headphones. At lO-sec intervals during adaptation, various probe tones were presented for 1 sec (the frequency of the probe was always the same as that of the adaptation stimulus). Observers judged the direction of apparent movement (“left” or “right”) of each probe tone. At 500 Hz, with a 200-deg/sec adaptation velocity, “stationary” probe tones were consistently judged to move in the direction opposite to that of the adaptation stimulus. We call this result an auditory motion aftereffect. In slower velocity adaptation conditions, progressively less aftereffect was demonstrated. In the higher frequency condition (2,000 Hz, 200-deg/sec adaptation velocity), we found no evidence of motion aftereffect. The data are discussed in relation to the well-known visual analog-the “waterfall effect.” Although the auditory aftereffect is weaker than the visual analog, the data suggest that auditory motion perception might be mediated, as is generally believed for the visual system, by direction-specific movement analyzers.