Orientation-invariant object recognition: evidence from repetition blindness

The question of whether object recognition is orientation-invariant or orientation-dependent was investigated using a repetition blindness (RB) paradigm. In RB, the second occurrence of a repeated stimulus is less likely to be reported, compared to the occurrence of a different stimulus, if it occurs within a short time of the first presentation. This failure is usually interpreted as a difficulty in assigning two separate episodic tokens to the same visual type. Thus, RB can provide useful information about which representations are treated as the same by the visual system. Two experiments tested whether RB occurs for repeated objects that were either in identical orientations, or differed by 30, 60, 90, or 180 degrees . Significant RB was found for all orientation differences, consistent with the existence of orientation-invariant object representations. However, under some circumstances, RB was reduced or even eliminated when the repeated object was rotated by 180 degrees , suggesting easier individuation of the repeated objects in this case. A third experiment confirmed that the upside-down orientation is processed more easily than other rotated orientations. The results indicate that, although object identity can be determined independently of orientation, orientation plays an important role in establishing distinct episodic representations of a repeated object, thus enabling one to report them as separate events.     

Effects of stimulus orientation on the identification of common polyoriented objects

Experimental evidence has shown that the time taken to recognize objects is often dependent on stimulus orientation in the image plane. This effect has been taken as evidence that recognition is mediated by orientation-specific stored representations of object shapes. However, the factors that determine the orientation specificity of these representations remain unclear. This issue is examined using a word-picture verification paradigm in which subjects identified line drawings of common mono- and polyoriented objects at different orientations. A detailed analysis of the results showed that, in contrast to mono-oriented objects, the recognition of polyoriented objects is not dependent on stimulus orientation. This interaction provides a further constraint on hypotheses about the factors that determine the apparent orientation specificity of stored shape representations. In particular, they support previous proposals that objects are encoded in stored representations at familiar stimulus orientations.     

Orientation congruency effects for familiar objects: coordinate transformations in object recognition

How do observers recognize objects after spatial transformations? Recent neurocomputational models have proposed that object recognition is based on coordinate transformations that align memory and stimulus representations. If the recognition of a misoriented object is achieved by adjusting a coordinate system (or reference frame), then recognition should be facilitated when the object is preceded by a different object in the same orientation. In the two experiments reported here, two objects were presented in brief masked displays that were in close temporal contiguity; the objects were in either congruent or incongruent picture-plane orientations. Results showed that naming accuracy was higher for congruent than for incongruent orientations. The congruency effect was independent of superordinate category membership (Experiment 1) and was found for objects with different main axes of elongation (Experiment 2). The results indicate congruency effects for common familiar objects even when they have dissimilar shapes. These findings are compatible with models in which object recognition is achieved by an adjustment of a perceptual coordinate system.     

Shape constancy: The effects of changing shape orientation and the effects of changing the position of focal features

Five experiments examined the time taken to judge that two consecutive elongated geometrical shapes had the same structure, irrespective of their orientation. Shape transformations either changed the orientation of the principal axis while maintaining the relative locations of focal features or maintained the orientation of the principal axis while changing the relative locations of focal features, or they changed both. Experiment 1 demonstrated that changes in the orientation of the principal axis were more detrimental to matching than were changes in the locations of the shape’s focal features. Indeed, the time taken to match same-orientation shapes was the same as that taken to match shapes that maintained the same position in the visual field. Further experiments showed that this result was not due to differential apparent motion in the transformation conditions, that it was not due to response bias, and that it generalized across shapes. However, the result was different when subjects could predict the location of the to-be-matched stimulus. In this case, performance was principally affected by the position of the focal feature of the shape and not by the shape’s orientation. It is suggested that the results reflect the efficiency with which subjects can construct matching representations for the stimuli When subjects cannot predict stimulus locations, they generate representations by describing shape structure relative to the shape’s principal axis. When the axis of the to-be-matched shapes is constant, subjects can use the same procedure in generating this representation for both shapes, facilitating matching relative to the case in which the orientation of the axis changes. When subjects can predict the stimulus location, they selectively attend to the focal features of shapes, minimizing the effects of shape orientation.     

Intrinsic array structure is neither necessary nor sufficient for nonegocentric coding of spatial layouts

Mou and McNamara (2002) have recently theorized that nonegocentric reference frames (viz., intrinsic reference frames, based on the spatial structure of a configuration of objects) are used to organize spatial relationships in memory. The theory has not made claims about whether the intrinsic structure of a stimulus array is necessary or sufficient for such nonegocentric coding. We demonstrate that salient intrinsic axes in a layout of objects are neither necessary nor sufficient for people to use a nonegocentric reference frame in organizing spatial memory. In Experiment 1, participants were successfully instructed to adopt a nonegocentric preferred direction in memory for an array of objects with no salient intrinsic reference axes. In Experiment 2, with no instructions, participants adopted an egocentric preferred direction for an array with a salient intrinsic axis. These results suggest that physically salient array structure exerts a minimal influence in the coding of spatial memory through nonegocentric reference frames.     

The analysis of orientation-dependent time costs in visual recognition

How does the visual system recognise stimuli presented at different orientations? According to the multiple-views hypothesis, misoriented objects are matched to one of several orientation-specific representations of the same objects stored in long-term memory. Much of the evidence for this hypothesis comes from the observation of group mean orientation effects in recognition memory tasks showing that the time taken to identify objects increases as a function of the angular distance between the orientation of the stimulus and its nearest familiar orientation The aim in this paper is to examine the validity of this interpretation of group mean orientation effects. In particular, it is argued that analyses based on group performance averages that appear consistent with the multiple-views hypothesis may, under certain circumstances, obscure a different theoretically relevant underlying pattern of results. This problem is examined by using hypothetical data and through the detailed analysis of the results from an experiment based on a recognition memory task used in several previous studies. Although a pattern of results that is consistent with the multiple-views hypothesis was observed in both the group mean performance and the underlying data, it is argued that the potential limitations of analyses based solely on group performance averages must be considered in future studies that use orientation effects to make inferences about the kinds of shape representations that mediate visual recognition.     

Orientation-specific effects in picture matching and naming

In two experiments, subjects made timed decisions about the second of two sequentially presented rotated drawings of objects. When the two objects were physically identical, response times to decide whether the two drawings depicted the same object varied as a function of the shortest distance between the orientation of the second drawing and either the orientation of the previous drawing or the upright. This was found for both short (250-msec) and long (2-sec) interstimulus-intervals. The result was also obtained when subjects named the second drawing after deciding whether the first drawing faced left or right. Following repeated experience with the drawings in the left/right task over four blocks of trials, time to name the second drawing in the same-object sequences was independent of orientation. These results suggest that, initially, object- and orientation-specific representations can be formed following a single presentation of a rotated object and subsequently used to identify drawings of the same object at either the same or different orientations. Alignment of the second drawing with either the canonical representation or the new representation at the previous orientation is achieved by normalization through the shortest path. Following experience with the objects, orientation-invariant representations are formed.     

Turning objects on their heads: the influence of the stored axis on object individuation

We used repetition blindness (RB) as a measure of object recognition and compared the pattern of RB obtained for objects with a well-established upright orientation (mono-oriented objects) and those without a usual upright orientation (polyoriented objects), when the critical objects were either in identical orientations or differed by 30 degrees, 60 degrees, 90 degrees, or 180 degrees. Overall, we found robust RB despite differences in orientation, consistent with the idea that object recognition, as indexed by RB, is largely independent of orientation. However, whereas for polyoriented objects RB was obtained in all orientation conditions, for mono-oriented objects there was no RB between upright and upside-down versions of the stimuli. These findings suggest that the usual orientation of an object-when it exists-is stored in memory and can facilitate orientation processing when the principal axis of a viewed object is aligned with the stored axis orientation. This, in turn, allows for more rapid and successful construction of distinct episodic representations of an object, thus alleviating RB.     

Orientation invariance in visual object priming depends on prime-target asynchrony

Two experiments are reported in which orientation effects on visual object recognition latency were examined. In Experiment 1, we assessed picture-naming performance as a function of image-plane stimulus orientation and found increasing response times with increased misorientation of the stimulus. In Experiment 2, we examined the repetition priming effect on the identification of upright targets as a function of prime orientation. With time delays of 100, 200, or 500 msec between the onset of the prime and that of the target (i.e., stimulus onset asynchrony [SOA]), the magnitude of the priming effect decreased with increasing misorientation of the prime. These results contrast with the orientation-invariant priming effects reported in some previous repetition priming studies. These investigations all used relatively long prime-target SOAs. Confirming the crucial role of the latter variable, Experiment 2 shows that the magnitude of the repetition priming effect is invariant across prime orientations with an SOA of 1,000 msec. The possible implications of the present observations with respect to the issue of orientation invariance versus dependency of the visual object recognition processare discussed.     

空间场景表征中的参照系选取

被试在矩形房间中从两个不同的观察点学习物体场景并在多个朝向上对物体形成的空间关系进行判断,通过控制场景中物体主要内在轴相对于环境结构（房间和地毯）的方向和被试的学习顺序,探讨被试在场景空间表征中采用何种参照系和参照系选取时的影响因素。两个实验结果发现：（1）内在参照系（intrinsic reference systems）和环境参照系均可以用于物体场景的表征,两类参照系之间的关系却是影响被试物体场景表征时参照系选取的重要因素,即当内在参照系与环境参照系方向一致时,被试无论从哪个朝向学习,都选择从垂直于内在参照系和环境参照系的朝向进行表征。反之,当二者方向不一致时,表征时参照系的选择取决于被试的学习经历;（2）无论内在参照系与环境参照系方向是否一致,物体场景本身内在结构的规则性都能够促进空间记忆,即内在结构的规则性既有助于准确编码物体的相对位置,也有助于提高空间关系判断的准确性。     

Orientation invariance in naming rotated objects: individual differences and repetition priming

In naming drawings of complex common objects, unpracticed naming times increase with rotation away from the upright, but this orientation effect is attenuated with practice. In principle, attenuation could result from learning to extract orientation-invariant information or from learning view-specific representations at the trained orientations. We contrasted these approaches by examining repetition priming for prime-target pairs presented on successive trials in either the same orientation (horse at 51 degrees primes horse at 51 degrees) or a different orientation (horse at 154 degrees primes horse at 51 degrees), for two subgroups of subjects. One subgroup showed no orientation effect, even when unpracticed, and a correspondingly high generalization of priming across different views. The other subgroup initially showed high sensitivity to misorientation and little priming across orientations but, with sufficient practice, came to show no orientation effect and complete generalization of priming. Thus, some subjects always used orientation-invariant procedures, whereas others learned to do so.     

Intrinsic orientation and study viewpoint in recognizing spatial structure of a shape

Two experiments dissociated the roles of intrinsic orientation of a shape and participants’ study viewpoint in shape recognition. In Experiment 1, participants learned shapes with a rectangular background that was oriented differently from their viewpoint, and then recognized target shapes, which were created by splitting study shapes along different intrinsic axes, at different views. Results showed that recognition was quicker when the study shapes were split along the axis parallel to the orientation of the rectangular background than when they were split along the axis parallel to participants’ viewpoint. In Experiment 2, participants learned shapes without the rectangular background. The results showed that recognition was quicker when the study shape was split along the axis parallel to participants’ viewpoint. In both experiments, recognition was quicker at the study view than at a novel view. An intrinsic model of object representation and recognition was proposed to explain these findings.     

The role of attention in the shift from orientation-dependent to orientation-invariant identification of disoriented objects

In two experiments, the naming of rotated line drawings of natural objects was examined after a training phase in which the objects were either attended or ignored. In the training phase of Experiment 1, subjects were presented with objects in a number of orientations over five repeated blocks of trials. In the center of each object, seven letters (Xs and Ts, colored red or blue) were presented in rapid succession. Half the subjects named aloud the rotated object and ignored the changing letter display (object-attend). The other half ignored the object and counted the number of red Ts, and then used this number to perform a simple multiplication (object-ignore). In the test phase, all subjects named the rotated objects. The results showed that in the first block of trials in the training phase, mean naming time in the object-attend condition increased the further an object was rotated from the upright. This effect of orientation for attended objects was much reduced in the later presentations of the test phase. In contrast, there was no such benefit of prior presentation observed for the naming of objects that had previously been ignored. Instead, a substantial orientation effect was shown for the naming of previously ignored objects, which was similar to the orientation effect observed for attended objects named in the first block. Similar results were found in Experiment 2, in which object-attend subjects in training covertly named the objects and then performed a letter count and multiplication task. In both experiments, performance on the letter count and multiplication task varied with the angle of the ignored object. The results suggest that full attentional resources must be allocated in order for orientation-invariant representations to be formed and used in the identification of rotated objects.     

Regularity of symmetry verticality guides perceptual judgments of objects

Previous research indicated that most salient, real-world objects possess natural regularities that observers commonly assume in perceptual judgments of figural orientation and interpretation. Regularities include 3-dimensionality, bilateral symmetry, and the tendency for object tops to possess more salient information than bottoms. Thus, when observers interpret randomly shaped figures, they reliably impose volume, bilateral symmetry, and top and front orientation directions, even when figures are 2-dimensional and asymmetric. We confirmed generalizability for observers to assume these regularities with stimuli that vary in complexity, and we found evidence supporting another regularity, that of symmetry verticality (symmetry about a vertical axis). Findings support use of a family of perceptual heuristics corresponding to natural regularities that constrain stimulus indeterminacy and help guide judgment of object orientation and interpretation.     

The role of axes of elongation and symmetry in rotated object naming

Many theorists have postulated that axes of elongation and/or symmetry play an important role in the recognition of objects. In this paper, evidence is presented that mitigates this claim from independent assessments of the effects of axes of elongation or symmetry on the time to name rotated line drawings of common objects. This conclusion was further supported in a stronger test in which both of these variables were orthogonally controlled, the aspect ratio of elongation was manipulated,and only objects that were completely geometrically symmetrical or asymmetrical were used. In all the experiments, objects were named for several blocks to determine the influence of these variables on effects of orientation with practice. Symmetry was found to diminish the effects of orientation after practice in naming the object set, and the effects of the most extreme orientation tested (120 degrees from upright) were diminished when both axes defined the same orientation, relative to when they defined different orientations. Contrary to many theories, these findings relegate the axes of symmetry and elongation to relatively minor roles during object identification.     

Roles of egocentric and allocentric spatial representations in locomotion and reorientation

Four experiments investigated the nature of spatial representations used in locomotion. Participants learned the layout of several objects and then pointed to the objects while blindfolded in 3 conditions: before turning (baseline), after turning to a new heading (updating), and after disorientation (disorientation). The internal consistency of pointing in the disorientation condition was relatively high and equivalent to that in the baseline and updating conditions, when the layout had salient intrinsic axes and the participants learned the locations of the objects on the periphery of the layout. The internal consistency of pointing was disrupted by disorientation when participants learned the locations of objects while standing amid them and the layout did not have salient intrinsic axes. It was also observed that many participants retrieved spatial relations after disorientation from the original learning heading. These results indicate that people form an allocentric representation of object-to-object spatial relations when they learn the layout of a novel environment and use that representation to locate objects around them. Egocentric representations may be used to locate objects when allocentric representations are not of high fidelity.     

Perceiving shape from profiles

Four experiments related human perception of shape from profiles to current theoretical predictions. In Experiment 1, judgments of structure and motion were obtained for single- and dualellipsoid displays rotating about various axes. Ratings were highest when the axis of rotation was in the image plane and were influenced by the number of ellipsoids and the orientation of a single ellipsoid. The subsequent experiments explored the effect of orientation on shape judgments of a single ellipsoid. The results of Experiments 2 and 3 suggested that the effect of orientation found in Experiment 1 was not due to either the inability of certain orientations to be perceived as three-dimensional objects or to two-dimensional artifacts. It was thus argued that this effect of orientation was due to points of correspondence in relative motion that arise when the major axis is not perpendicular to the axis of rotation. In Experiment 4, subjects provided judgments of both shape and angular velocity. The elevated ellipsoids that were judged as larger were also judged as rotating more slowly. The inverse relationship between size and angular velocity is consistent with current theories. The connection between theory and data was further demonstrated by applying a shape-recovery algorithm to the stimuli used in Experiment 4 and finding a similar tradeoff between angular velocity and shape.