Orientation tuning of shape from shading

Four experiments were performed to assess the effect of different orientations and direction of lighting on the visual processing of shaded or bipartite disks. In the first two experiments, observers were presented with nine different shading orientations from 0 degree to 180 degrees. Targets were detected in a rapid and parallel fashion for shaded disks when the orientation of the shading gradient was not horizontal (90 degrees) or oriented at 67.5 degrees. Search asymmetries favoring the detection of "pock" targets over "ball" targets were found for all orientations. The search rates for bipartite disks were similar to the shaded disks at 0 degree, 22.5 degrees, and 90 degrees but not for intermediate orientations, and no search asymmetries were found. These differences suggest that shaded displays and bipartite displays are processed by different underlying mechanisms. The third experiment showed that the direction of the light source (left or right) had no influence on search asymmetries around the 90 degrees point. Shading gradient orientation affected magnitude estimates of depth in the fourth experiment. These experiments show that the visual system's "assumption" of overhead lighting is broadly tuned.     

Asymmetry in the detection of shapes from shading in infants1

Abstract: We investigated 3‐ and 4‐month‐old infants’ sensitivity to differences defined by shading using a paired‐comparison familiarity/novelty preference procedure. Infants were familiarized with a pair of displays consisting of homogeneous shaded disks, and then were tested with two displays: the familiar display and a novel one containing shaded disks with reversed polarity (defined as the target). Experiment 1 examined two assumptions on discerning shapes from shading in infants by manipulating the orientations in the shading gradient of stimuli. When the orientation of the shading gradient was vertical, 4‐month‐old infants looked at the novel display for a longer time during the test trial. However, they failed to detect differences when the orientation of shading gradients was horizontal. Three‐month‐old infants did not detect differences in either orientation of the shading gradient. Experiment 2 examined asymmetry in the detection of convex versus concave shapes. Four‐month‐old infants failed to detect the target when the orientation of the shading grating was vertical and the target was convex. Taken with the results of Experiment 1, concave shapes were much easier to detect than convex shapes for 4‐month‐olds. This asymmetry suggests that 4‐month‐old infants process shading information in the same manner as adults.     

Visual search for direction of shading is influenced by apparent depth

Recent reports ofrapid visual search for some feature conjunctions-suggested that preattentive vision might be sensitive to scene-based as well as to image-based features (Enns & Rensink, 1990a, 1990b). This study examined visual search for targets defined by the direction of a luminance gradient, a conjunction of luminance and relative location that often: corresponds to ob-= ject curvature and direction of lighting in naturalistic scenes. Experiment 1 showed that such search is influenced by several factors, including the type of gradient, the shape of the contour enclosing the gradient, and the background luminance. These factors were varied systematically in Experiment 2 in a three-dimensionality rating task and in a visual-search task. The factors combined interactively in the rating task, supporting the presence of an emergent property of three-dimensionality. In contrast, each factor contributed only additively to the speed of the visualsearch task. This is inconsistent with the view that search is guided by specialized detectors for surface curvature or direction of lighting. Rather, it is in keeping with the view that search is governed by a number of “quick and dirty” processes that are implemented rapidly and in parallel across the visual field.     

On the perception of shape from shading

The extraction of three-dimensional shape from shading is one of the most perceptually compelling, yet poorly understood, aspects of visual perception. In this paper, we report several new experiments on the manner in which the perception of shape from shading interacts with other visual processes such as perceptual grouping, preattentive search (“pop-out”), and motion perception. Our specific findings are as follows: (1) The extraction of shape from shading information incorporates at least two “assumptions” or constraints—first,that there is a single light source illuminating the whole scene, and second, that the light is shining from “above” in relation to retinal coordinates. (2) Tokens defined by shading can serve as a basis for perceptual grouping and segregation. (3) Reaction time for detecting a single convex shape does not increase with the number of items in the display. This “pop-out” effect must be based on shading rather than on differences in luminance polarity, since neither left-right differences nor step changes in luminance resulted in pop-out. (4) When the subjects were experienced, there were no search asymmetries for convex as opposed to concave tokens, but when the subjects were naive, cavities were much easier to detect than convex shapes. (5) The extraction of shape from shading can also provide an input to motion perception. And finally, (6) the assumption of “overhead illumination” that leads to perceptual grouping depends primarily on retinal rather than on “phenomenal” or gravitational coordinates. Taken collectively, these findings imply that the extraction of shape from shading is an “early” visual process that occurs prior to perceptual grouping, motion perception, and vestibular (as well as “cognitive”) correction for head tilt. Hence, there may be neural elements very early in visual processing that are specialized for the extraction of shape from shading.     

Target-tilt and vertical-hemifield asymmetries in free-scan search for 3-D targets

In this study, asymmetries in finding pictorial 3-D targets defined by their tilt and rotation in space were investigated by means of a free-scan search task. In Experiment 1, feature search for cube tilt and rotation, as assessed by a spatial forced-choice task, was slow but still exhibited a characteristic “flat” slope; it was also much faster to upward-tilted cubes and to targets located in the upper half of the search field. Faster search times for cubes and rectangular solids in the upper field, an advantage for upward-tilted cubes, and a strong interaction between target tilt and direction of lighting (upward or downward) for the rectangular solids were all demonstrated in Experiment 2. Finally, an advantage in searching for tilted cubes located in the upper half of the display was shown in Experiment 3, which used a present-absent search task. The results of this study confirm that the upper-field bias in visual search is due mainly to a biased search mechanism and not to the features of the target stimulus or to specific ecological factors.     

Learning to search for visual targets defined by edges or by shading: Evidence for non-equivalence of line drawings and surface representations

A visual search task was used to test the idea that shaded images and their line-drawn analogues are treated identically from an early stage onwards in human vision. Reaction times and error rates were measured to locate the presence or absence of a target in an array of a variable number of distractors. The target was a cube in one orientation and the distractors cubes in a different orientation. The stimuli were defined by lines alone, shading alone, or lines plus shading. Both the slopes and the intercepts of the search functions (graphs of search time against number of displayed items) were higher for the line drawings than for the stimuli defined by shading. Over six experimental sessions, both the slopes and the intercepts fell for all stimuli, but the relative differences between them were maintained. The data suggest that, at an equivalent stage of practice, line-drawn stimuli are processed more slowly than shaded stimuli in early vision.     

Visual search for simple volumetric shapes

Five experiments measured reaction time (RT) to detect the presence or absence of a simple volumetric shape (target) dependent on the number of display items (distractors) and their depicted three-dimensional (3-D) orientation. Experiments 1–4 examined every pairwise combination of two different simple volumetric shapes in two 3-D orientations. Conditions exhibiting “pop-out” could be predicted by differences in their two-dimensional (2-D) features. Conditions in which search was slower support previously found search asymmetries for particular 2-D features. When the distractors were a mixture of the other shapes in the same 3-D orientation, search was serial, except when the target had a curved principal axis (Experiment 5). The results suggest that these simple volumetric shapes are not processed preattentively.     

The discrimination of structure in vectorgraphs: Local and global effects

Many recent electrophysiological studies have demonstrated specificity for orientation and position tuning in single units of the visual system. Psychophysical investigations have produced complementary evidence for orientation and position coding in vision which is usually interpreted in terms of the underlying neuronal properties. To date, no studies have endeavored to relate such “local” receptive field response profiles to more “global” or field-specific aspects of the input signal. In a series of experiments, we have investigated orientation/position sensitivity as a function of general orientation and position rules governing line element, or “vectorgraph,” displays. We have discovered with such vectorgraph images that individual element acuities are determined by the field structures present. These results show that there are specific “top-down” components even to basic orientation/position coding in the visual system     

Spatial frequency comparisons of rotated gratings

Reaction time for determining if two square-wave gratings are the same or different is a bimodal function of the relative angular separation (RAS) between them. Response time peaks at 60° and 120° separation and drops siguificantly at approximately 90°, provided that the gratings being compared are of the same spatial frequency. This pattern of results is not in line with those previously reported for more complex figures, nor may it be attributed to any “oblique effect,” since the same function is obtained with the comparison gratings placed at different retinal orientations. One possible explanation is that orthogonally tuned orientation channels interact in a facilitatory fashion in the human visual system.     

Monkeys perceive the orientation of objects relative to the vertical axis

Orientation processing is essential for segmenting contour from the background, which allows perception of the shape and stability of objects. However, little is known about how monkeys determine the degree and direction of orientation. In this study, to determine the reference axis for orientation perception in monkeys, post-discrimination generalization tests were conducted following discrimination training between the 67.5° and 112.5° orientations and between the 22.5° and 157.5° orientations. After discrimination training between the 67.5° and 112.5° orientations, the slope of the generalization gradient around the S⁺ orientation was broad, while the slope was steep after discrimination training between the 22.5° and 157.5° orientations. Comparing the shapes of the gradients indicated that the subjective distance between the 67.5° and 112.5° orientations was small, while the subjective distance between the 22.5° and 157.5° orientation was large. In other words, the monkeys recognized that the former and the latter distances were 45° and 135° across the vertical axis, rather than 135° and 45° across the horizontal axis, respectively. These findings indicate that the monkeys determined the degree and direction of the tilt using the vertical reference.     

Shape from shading in pigeons

Light is the origin of vision. The pattern of shading reflected from object surfaces is one of several optical features that provide fundamental information about shape and surface orientation. To understand how surface and object shading is processed by birds, six pigeons were tested with differentially illuminated convex and concave curved surfaces in five experiments using a go/no-go procedure. We found that pigeons rapidly learned this type of visual discrimination independent of lighting direction, surface coloration and camera perspective. Subsequent experiments varying the pattern of the lighting on these surfaces through changes in camera perspective, surface height, contrast, material specularity, surface shape, light motion, and perspective movement were consistent with the hypothesis that the pigeons were perceiving these illuminated surfaces as three-dimensional surfaces containing curved shapes. The results suggest that the use of relative shading for objects in a visual scene creates highly salient features for shape processing in birds.     

Determining the orientation of depth-rotated familiar objects

How does the human visual system determine the depth-orientation of familiar objects? We examined reaction times and errors in the detection of 15° differences in the depth orientations of two simultaneously presented familiar objects, which were the same objects (Experiment 1) or different objects (Experiment 2). Detection of orientation differences was best for 0° (front) and 180° (back), while 45° and 135° yielded poorer results, and 90° (side) showed intermediate results, suggesting that the visual system is tuned for front, side and back orientations. We further found that those advantages are due to orientation-specific features such as horizontal linear contours and symmetry, since the 90° advantage was absent for objects with curvilinear contours, and asymmetric object diminished the 0° and 180° advantages. We conclude that the efficiency of visually determining object orientation is highly orientation-dependent, and object orientation may be perceived in favor of front-back axes.     

Similarity: Its definition and effect on the visual analysis of complex displays

Contemporary feature models of form perception have typically defined visual similarity in terms of shared (or discordant) sets of points. Two experiments tested the adequacy of this definition. In a same/different task, subjects were required to detect a single “different” form in displays of two, four, or six forms. In separate conditions, the “different” form was produced by various geometric transformations, where the number of discordant points could be held constant for some of those transformations. The first experiment compared the detectability of three transformations: deletion of an end-of-a-line segment, a break in continuity, and a mirror-image reversal. Reversals were detected most rapidly and accurately, with performance independent of display size. Although breaks and deletions produced the same number of discordant points, breaks were detected more rapidly and accurately. The second experiment tested whether the better detectability of reversals was due to a greater number of discordant points or to changes in the orientation of diagonal lines. The results indicated that entire displays can be rapidly organized (in “parallel”) on the basis of line orientations. In general, the experiments suggest that the similarity of forms may depend upon the transformations by which they are related rather than their common features.     

Visual search for direction of shading is influenced by apparent depth.

Recent reports of rapid visual search for some feature conjunctions suggested that preattentive vision might be sensitive to scene-based as well as to image-based features (Enns & Rensink, 1990a, 1990b). This study examined visual search for targets defined by the direction of a luminance gradient, a conjunction of luminance and relative location that often corresponds to object curvature and direction of lighting in naturalistic scenes. Experiment 1 showed that such search is influenced by several factors, including the type of gradient, the shape of the contour enclosing the gradient, and the background luminance. These factors were varied systematically in Experiment 2 in a three-dimensionality rating task and in a visual-search task. The factors combined interactively in the rating task, supporting the presence of an emergent property of three-dimensionality. In contrast, each factor contributed only additively to the speed of the visual-search task. This is inconsistent with the view that search is guided by specialized detectors for surface curvature or direction of lighting. Rather, it is in keeping with the view that search is governed by a number of "quick and dirty" processes that are implemented rapidly and in parallel across the visual field.     

The associated nontargets effect: Evidence of nontarget processing at and beyond the level of letter recognition

In four visual search tasks participants were asked to make a target response if either of two targets was present and to make a nontarget response if neither target was present. Some target-absent displays included only nontarget stimuli or features that never occurred in the same displays as targets, whereas other target-absent displays included nontarget stimuli or features that did sometimes occur with targets. Nontarget responses were reliably faster in the former case than in the latter. This “associated nontargets effect” indicates that nontargets are not simply classified as nontargets but in addition are discriminated from one another. Current visual search models may underestimate the degree to which nontargets are processed during search.     

Orientation categories used in guidance of attention in visual search can differ in strength

We investigated orientation categories in the guidance of attention in visual search. In the first two experiments, participants had a limited amount of time to find a target line among distractors lines. We systematically varied the orientation of the target and the angular difference between the target and distractors. We find vertical, horizontal, and 45° targets require the least target/distractor angular difference to be found reliably and that the rate at which increases in target/distractor difference decrease search difficulty to be independent of target identity. Unexpectedly, even when the angular difference between the target and distractors was large, search performance was never optimal when the target orientation was 45°. A third experiment investigates this unexpected finding by correlating target/distractor difference and error rate with performance on tasks that measure a specific perceptual or cognitive ability. We find that the elevated error rate is correlated with performance on stimulus recognition and identification tasks, while the amount of target/distractor difference needed to detect the target reliably is correlated with performance on a stimulus reproduction task. We conclude that the target/distractor difference reveals the number of orientation categories in visual search, and, accordingly, that there are four such categories: two strong ones centred on 0° and 90° and two weak ones centred on 45° and 135°.     

Spatio-temporal probing of apparent rotational movement

A visual bar alternately presented in vertical and horizontal orientations appeared to rotate 90° through one of two pairs of opposite quadrants. Subjects judged whether a probe dot, interjected at some delay and angular deviation from the vertical bar, appeared before or after the bar “passed through” the corresponding angular orientation. When the motion was perceived in the probed quadrant, percent “before” responses dropped abruptly from near 100% to near 0% as delay increased, and the drop occurred at longer delays for probes at larger angular deviations. Under physically identical conditions, when the motion was perceived in an unprobed quadrant, percent “before” responses varied much less with delay and insignificantly with angular position. The accuracy of the judgments in the first case suggests the internal generation of an ordered sequence of intermediate representations during each apparent rotation.     

How apparent motion affects mental rotation: Push or pull?

Subjects were timed as they judged whether a small bar perpendicular to one side of a clockhand would point left or right if the hand was pointing upward (i.e., at the 12:00 position). The clockhand was shown in two successive orientations 30°apart, so that it was perceived to jump from one to the other, but the bar was included at only one of the two orientations. Analysis of reaction times as a function of angular orientation showed that the subjects “mentally rotated” the clockhand to the upright position before making their decisions. When the bar appeared on the second presentation, the jump had no significant influence on mental rotation but when it appeared on the first presentation, the estimated orientation from which the clockhand was mentally rotated was “dragged” in the direction of the jump.     

Guided search through memory

In “hybrid” search, observers search a visual display for any of several targets held in memory. It is known that the contents of the memory set can guide visual search (e.g., if the memorized targets are all animals, visual attention can be guided away from signs). It is not known if the visual display can guide memory search (e.g., if the memory set is composed of signs and animals, can a visual display of signs restrict memory search to just the signs?). In three hybrid search experiments, participants memorized sets of items that belonged to either one or several categories. Participants were then presented with visual displays containing multiple items, also drawn from one or several categories. Participants were asked to determine if any of the items from their current memory set were present in the visual display. We replicate the finding that visual search can be guided by the contents of memory. We find weaker, novel evidence that memory search can be guided by the contents of the visual display.     

Feature analysis and the role of similarity in preattentive vision

Texture arrays of line elements at various orientations were used to study three phenomena of preattentive vision. Subjects were asked (1) to discriminate texture areas and to distinguish their form (experiments on texture segmentation); (2) to detect salient or vertical line elements (experiments on pop-out); and (3) to identify configurations of similar or dissimilar targets (experiments on grouping). Within the patterns, line orientation was systematically varied to distinguish the effect of differences between areas from the effect of similarity within areas. In all of the experiments, performance was found to depend on local orientation contrast at texture borders rather than on the analysis of line orientation itself. Texture areas were correctly identified only when the orientation contrast at the border well exceeded the overall variation of line orientation in the pattern. Similarly, only target elements with high local orientation contrast were detected fast and “in parallel.” Targets with an orientation contrast lower than background variation required serial search. Preattentive grouping was found to depend on saliency, as defined by local orientation contrast, but not on the similarity of line elements. In addition to local orientation contrast, which played an important role in all of the visual phenomena studied, influences from the alignment of line elements with the outline of a figure were also seen.