Visual Search Asymmetry Research Articles

Performance asymmetries in visual search demonstrate failure of independent-processing models

We report psychophysical data from orientation-popout experiments that are inconsistent with a rather general decision model. Stimuli consisted of 121 line segments arranged on an 11×11 grid. There were two tasks: in the 1-Singleton Task all lines except one had the same orientation, and observers had to report which quadrant contained the singleton. In the 3-Singleton Task three quadrants contained orientation singletons and observers had to identify the quadrant without singleton. These tasks can be viewed as asymmetric search tasks, in which either a singleton-quadrant has to be found among three homogeneous quadrants, or a homogeneous quadrant has to be found among three singleton-quadrants. Using tools from signal-detection theory we show that the large performance asymmetries between 1-Singleton and 3-Singleton Tasks are inconsistent with any model that makes two (very basic and common) assumptions: (1) independent processing of the four quadrants and (2) an ideal-observer decision. We conclude that at least one of the two assumptions is inadequate. As a plausible reason for the model failure we suggest a global competition between salient elements that reduces popout strength when more than one singleton is present.

Contextual influences in V1 as a basis for pop out and asymmetry in visual search.

I use a model to show how simple, bottom-up, neural mechanisms in primary visual cortex can qualitatively explain the preattentive component of complex psychophysical phenomena of visual search for a target among distracters. Depending on the image features, the speed of search ranges from fast, when a target pops-out or is instantaneously detectable, to very slow, and it can be asymmetric with respect to switches between the target and distracter objects. It has been unclear which neural mechanisms or even cortical areas control the ease of search, and no physiological correlate has been found for search asymmetry. My model suggests that contextual influences in V1 play a significant role.

Visual texture segregation by the chimpanzee ( Pan troglodytes)

One adult male chimpanzee ( Pan troglodytes) was trained to detect a target area consisting of texture elements from the background texture made of the different elements from the target area. The subject was given eight different stimulus conditions. In Condition 1, segregation was based on the difference of local feature of elements but not on global similarity. Conditions 2–3 investigated the effects of the number of terminators, which was considered as `textons' in human texture perception. The chimpanzee showed better performance when the discrimination can be based on the difference of the number of terminators. This tendency, however, was reversed when the local salient feature (length of shorter lines) was enhanced, as in Enns' [15]study with humans as subjects. The subject showed asymmetries in segregation performance when discriminating based on gap (Conditions 4–5), line length (Condition 6), and regularity of line arrangement (Condition 7). Observed asymmetries were consistent with humans and with visual search asymmetries. The performance of texture segregation by the chimpanzee was consistent with humans, and the texture segregation is one of the useful tasks for comparative study of early vision as well as visual search task.

Visual segmentation by contextual influences via intra-cortical interactions in the primary visual cortex

Stimuli outside classical receptive fields have been shown to exert a significant influence over the activities of neurons in the primary visual cortex. We propose that contextual influences are used for pre-attentive visual segmentation. The difference between contextual influences near and far from region boundaries makes neural activities near region boundaries higher than elsewhere, making boundaries more salient for perceptual pop-out. The cortex thus computes global region boundaries by detecting the breakdown of homogeneity or translation invariance in the input, using local intra-cortical interactions mediated by the horizontal connections. This proposal is implemented in a biologically based model of V1, and demonstrated using examples of texture segmentation and figure-ground segregation. The model is also the first that performs texture or region segmentation in exactly the same neural circuit that solves the dual problem of the enhancement of contours, as is suggested by experimental observations. The computational framework in this model is simpler than previous approaches, making it implementable by V1 mechanisms, though higher-level visual mechanisms are needed to refine its output. However, it easily handles a class of segmentation problems that are known to be tricky. Its behaviour is compared with psycho-physical and physiological data on segmentation, contour enhancement, contextual influences and other phenomena such as asymmetry in visual search.

Feature asymmetries in visual search: effects of display duration, target eccentricity, orientation and spatial frequency.

In Experiments 1-3, we monitored search performance as a function of target eccentricity under display durations that either allowed or precluded eye movements. The display was present either until observers responded, for 104 msec, or for 62 msec. In all three experiments an orientation asymmetry emerged: observers detected a tilted target among vertical distracters more efficiently than a vertical target among vertical distracters. As target eccentricity increased, reaction times and errors augmented, and the set size effect became more pronounced, more so for vertical than tilted targets. In Experiments 4-7, the stimulus spatial properties were manipulated: spatial frequency; size; and orientation. The eccentricity effect was more pronounced for vertical than tilted targets and for high- than low-spatial frequency targets. This effect was eliminated when either the size, the size and orientation, or the size and spatial frequency were magnified (M-cortical factor). By increasing the signal-to-noise ratio, magnification reduced the extent of both asymmetries; it aided more the detection of tilted than vertical and of high- than low-spatial frequency targets. Experiments 4-7 indicate that performance improvement in the magnified conditions was due to the specific pairing of stimulus size with retinal eccentricity and not to the larger stimulus size of the magnified conditions. We conclude that stimulus size, orientation and spatial frequency influence the extent of the eccentricity effect and the efficiency of search performance.

Electrophysiological correlates of visual search asymmetry in humans.

Visual evoked potentials (VEPs) were recorded during a visual search task to evaluate parallel and serial models of visual processing. A target stimulus elicited a discrete posterior negative wave in the 150-300 ms latency range. Furthermore, there was a close correlation between the search performance and the activity of the negative wave. When the target had a unique feature, neither the search time nor the characteristics of the negative wave was affected by the number of items in display. However, when the target lacked a unique feature, both the search time and the latency of the negative wave increased with the number of items. These results were consistent with a claim that an object with a unique feature is detected preattentively.

Visual search asymmetry for viewing direction.

In visual search experiments, we examined the existence of a search asymmetry for the direction with which three-dimensional objects are viewed. It was found that an upward-tilted target object among downward-tilted distracting objects was detected faster than when the orientation of target and distractors was reversed. This indicates that the early visual process regards objects tilted downward with respect to the observer as the situation that is more likely to be encountered. That is, the system is set up to expect to see the tops of these objects. We also found a visual field anisotropy, in that the asymmetry was more pronounced in the lower visual field. These findings are consistent with the idea that the tops of objects are usually situated in the lower visual field and less often in the upper field. Examination of the conditions under which the asymmetry and the anisotropy occur demonstrated the importance of the three-dimensional nature of the stimulus objects. Early visual processing thus makes use of heuristics that take into account specific relationships between the relative locations in space of the observer and 3-D objects.

Perceptual integration of motion and form information: is the movement filter involved in form discrimination?

Seven experiments reinvestigated J. Driver and P. McLeod's (1992) report of a "visual search asymmetry reversal" in a task that required the integration of motion and form information. They found that, when the form discrimination was easy, search was more efficient for a moving rather than a stationary conjunction target; the reverse was true when form discrimination was difficult. J. Driver and P. McLeod proposed that 2 mechanisms are involved: a "stationary form system," which supports accurate form discrimination but is relatively insensitive to movement, and a "movement filter," which segregates the moving from the stationary items but is relatively insensitive to aspects of form. The present experiments failed to find the asymmetry reversal. The results agree with the (more parsimonious) proposal that the function of the movement filter is limited to separating moving from stationary items, whereas form discrimination is accomplished within a unitary form system.

Visual search asymmetries in three-dimensional space

Feature-conjunction search performance was investigated as a function of the target's location in three-dimensional (3-D) space. Ten subjects viewed a display that contained 36 shapes, one of which was the target. The targets were presented in one of four quadrants, three depths, and three eccentricities. On a given trial, nontarget distractor shapes were presented at the other 35 locations. The ability to find a target shape was best when it was presented in the upper and right visual fields and was closest to the fixation point in both its depth and eccentricity. These nonuniformities may be partly derived from the link between visual search and an extrapersonal attentional mechanism.

Asymmetries in visual search for conjunctive targets.

Asymmetry is demonstrated between conjunctive targets in visual search with no detectable asymmetries between the individual features that compose these targets. Experiment 1 demonstrated this phenomenon for targets composed of color and shape. Experiment 2 and 4 demonstrate this asymmetry for targets composed of size and orientation and for targets composed of contrast level and orientation, respectively. Experiment 3 demonstrates that search rate of individual features cannot predict search rate for conjunctive targets. These results demonstrate the need for 2 levels of representations: one of features and one of conjunction of features. A model related to the modified feature integration theory is proposed to account for these results. The proposed model and other models of visual search are discussed.

Asymmetries in visual search for conjunctive targets.

Asymmetries in visual search for conjunctive targets.

Asymmetry in visual search for targets defined by differences in movement speed.

Asymmetry in visual search for targets defined by differences in movement speed.

Asymmetry in visual search for targets defined by differences in movement speed.

Perception of motion speed was investigated with the visual search paradigm, using human Ss. When searching for a fast target among slow distractors, reaction time was minimally affected as the number of distractors was increased. In contrast, reaction time to detect a slow target among fast distractors was slow and linearly related to the number of distractors. The effect cannot be attributed to differences in temporal frequency, discriminability, or one type of representation that might result from spatiotemporal filtering. An alternative hypothesis that can account for the asymmetry is that speed detectors operate as high-pass filters in the velocity domain. This hypothesis is in agreement with results obtained in psychophysical studies on motion adaptation as well as data from single-cell recordings in nonhuman species.

Reversing visual search asymmetries with conjunctions of movement and orientation.

Reversing visual search asymmetries with conjunctions of movement and orientation.

Reversing visual search asymmetries with conjunctions of movement and orientation.

Many theories of visual search offer general principles intended to apply to search for targets defined by any stimulus attribute. The importance of considering specific properties of different visual subsystems is demonstrated by examining search for movement-orientation conjunctions