Parts of recognition

Visual Cognition, 2003

In this study a change-detection paradigm is used to explore the nature of the information used to recognise three-dimensional, novel objects. In particular, whether we are sensitive to changes in part identity or configuration information. Experiments 1 and 2 showed that configural changes made to the object parts were significantly easier and quicker to detect than changes made to the shape or arrangement of object parts. Variance due to the total change in pixels did not predict performance in either of the experiments. The results of Experiment 3 showed the same pattern as Experiments 1 and 2, even though the objects used were altered to make part identity information more salient. Experiment 4 demonstrated that the physical size of the changes is not the crucial variable for this pattern of results. These findings are discussed in relation to the nature of visual representations and theories of object recognition.

Visual Cognition, 2001

It has been previously demonstrated that visual attention has an extent in depth (3D space) as well as an extent in the fronto-parallel plane (2D space). Numerous experiments have also demonstrated that attention can be allocated to objects, and that “object-based” attention can overcome some of the costs associated with moving attention about in 2D space. In real visual environments, objects often have an extent in depth. Four experiments were conducted to examine the nature of object-based attention in 3D space. The experiments demonstrated large object-based attention benefits, as well as costs for switching attention in depth. However, the costs associated with switching attention in depth were eliminated with objects that had an extent in depth. Experiments 2-4 examined the interaction of spatial attention in 3D space and object-based attention. Evidence was found for the spread of spatial attention to objects. However, contrary to other work (Lavie & Driver, 1996), neither non-predictive exogenous spatial cues (Experiment 2) nor predictive exogenous spatial cues (Experiments 3 and 4) were able to eliminate object-based attention, suggesting that object-based attention can remain intact despite the allocation of attention spatially.

Memory & Cognition, 2006

Everyday events, such as making a bed, can be segmented hierarchically, with the coarse level characterized by changes in the actor’s goals and the fine level by subgoals (Zacks, Tversky, & Iyer, 2001). Does hierarchical event perception depend on knowledge of actors’ intentions? This question was addressed by asking participants to segment films of abstract, schematic events. Films were novel or familiarized, viewed forward or backward, and simultaneously described or not. The participants interpreted familiar films as more intentional than novel films and forward films as more intentional than backward films. Regardless of experience and film direction, however, the participants identified similar event boundaries and organized them hierarchically. An analysis of the movements in each frame revealed that event segments corresponded to bursts of change in movement features, with greater bursts for coarse than for fine units. Perceiving event structure appears to enable event schemas, rather than resulting from them.

International Journal of Intelligent Systems, 1994

This article addresses the problem of the recognition of the usefulness of man-made objects in actions. In a recent article, [L. M. Vaina and M. C. Jaulent, Intelligent System Journal6, 313-336 (1991)] we defined the functional recognition problem in terms of the compatibility between the primary functions of objects and action requirements. We proposed a model, based on fuzzy sets and possibility theory, for the organization of this knowledge and the evaluation of the compatibility to perform functional recognition. In this article, we present an implementation of functional compatibility which takes the functional descriptions of objects and the object requirements of an action as inputs, and maps them into an output space that evaluates their functional compatibility. © 1994 John Wiley & Sons, Inc.

Attention Perception & Psychophysics, 2005

Coherent visual experience of dynamic scenes requires not only that the visual system segment scenes into component objects but that these object representations persist, so that an object can be identified as the same object from an earlier time. Object files (OFs) are visual representations thought to mediate such abilities: OFs lie between lower level sensory processing and higher level recognition, and they track salient objects over time and motion. OFs have traditionally been studied viaobjectspecific preview benefits (OSPBs), in which discriminations of an object’s features are speeded when an earlier preview of those features occurred on the same object, as opposed to on a different object, beyond general displaywide priming. Despite its popularity, many fundamental aspects of the OF framework remain unexplored. For example, although OFs are thought to be involved primarily in online visual processing, we do not know how long such representations persist; previous studies found OSPBs for up to 1,500msec but did not test for longer durations. We explored this issue using a modifiedobject reviewing paradigm and found that robust OSPBs persist for more than five times longer than has previously been tested—for at least 8 sec, and possibly for much longer. Object files may be the “glue” that makes visual experience coherent not just in online moment-by-moment processing, but on the scale of seconds that characterizes our everyday perceptual experiences. These findings also bear on research in infant cognition, where OFs are thought to explain infants’ abilities to track and enumerate small sets of objects over longer durations.

Ecological Psychology, 1991

Attention Perception & Psychophysics, 2007

Using a change detection paradigm, Barenholtz, Cohen, Feldman, and Singh (2003) found that changes in concave regions of a contour are more easily detected than changes in convex regions. In a series of three experiments, we investigated this concavity effect using the same paradigm. We observed the effect in wirelike stimuli as well as in silhouettes (Experiment 1) and in complex, smoothed images as opposed to angular polygons (Experiment 2). We also observed a systematic effect of the magnitude of the change (Experiment 1). Furthermore, we find that the effect cannot be attributed to either local or global processing effects, but rather to a combination of both “mere” concaveness and an effect due to changes in the perceived part structure of the stimulus object (Experiment 3). For our data analysis, we used a nonparametric bootstrap method, which greatly increases sensitivity (compared to more traditional analyses like ANOVA).

Perception, 2008

Although Attneave (1954, Psychological Review 61 183 -193) and Biederman (1987, Psychological Review 94 115 -147) have argued that curved contour segments are most important in shape perception, Kennedy and Domander (1985, Perception 14 367 -370) showed that fragmented object contours are more identifiable when straight segments are shown. Based on the set of line drawings published by Snodgrass and Vanderwart (1980, Journal of Experimental Psychology: Human Learning and Memory 6 174 -215), we have made outline versions that can be used to investigate this issue with a larger and more heterogeneous stimulus set.

2001

Human vision organizes object shapes in terms of parts and their spatial relationships.

Neurogastroenterology and Motility, 2001

1 Abstract The role of vagal innervation on emptying patterns and intragastric distributions of liquid and semisolid meals is still controversial. We aimed to record these features after dorsal, ventral and truncal vagotomies, using external gamma scintigraphy in conscious pigs in which the dorsal vagus speci®cally innervates the proximal stomach. Imaging of the stomach was performed for all experimental situations and before surgery using 99m Tc-labelled glucose and porridge meals. 2

Perception & Psychophysics, 2001

Studies of object-based attention have demonstrated poorer performance in dividing attention between two objects in a scene than in focusing attention on a single object. However, objects often are composed of several parts, and parts are central to theories of object recognition. Are parts also important for visual attention? That is, can attention be limited in the number of parts processed simultaneously? We addressed this question in four experiments. In Experiments 1 and 2, participants reported two attributes that appeared on the same part or on different parts of a single multipart object. Participants were more accurate in reporting the attributes on the same part than attributes on different parts. This part-based effect was not influenced by the spatial distance between the parts, ruling out a simple spatial attention interpretation of our results. A control study demonstrated that our spatial manipulation was sufficient to observe shifts of spatial attention. This study revealed an effect of spatial distance, indicating that our spatial manipulation was adequate for observing spatial attention. The absence of a distance effect in Experiments 1 and 2 suggests that part-based attention may not rely entirely on simple shifts of spatial attention. Finally, in Experiment 4 we found evidence for part-based attention, using stimuli controlled for the distance between the parts of an object. The results of these experiments indicate that visual attention can selectively process the parts of an object. We discuss the relationship between parts and objects and the locus of part-based attentional selection.

Psychonomic Bulletin & Review, 2000

Recent empirical results suggest that there is a decrement in dividing attention between two objects in a scene compared with focusing attention on a single object. However, objects can be made of individual parts. Is there a decrement for dividing attention across different parts of a single object? We addressed this question in two experiments. In Experiment 1, we demonstrated that attention can exhibit part-based selection-that is, the subjects were more accurate in reporting two attributes from the same part of an object than they were in reporting attributes from different parts of an object. In Experiment 2, we demonstrated that part-based attentional decrements occurred simultaneously with object-based attentional decrements. The results from Experiment 2 demonstrated that part-based attention is evident at the same time as objects are processed as coherent wholes. Our results imply that there is an attentional mechanism that can select either objects or their parts.

Perception & Psychophysics, 2002

The visual system represents object shapes in terms of intermediate-level parts. The minima rule proposes that the visual system uses negative minima of curvature to define boundaries between parts. We used visual search to test whether part structures consistent with the minima rule are computed preattentively-or at least, rapidly and early in visual processing. The results of Experiments 1 and 2 showed that whereas the search for a non-minima-segmented shape is fast and efficient among minimasegmented shapes, the reverse search is slow and inefficient. This asymmetry is expected if parsing at negative minima occurs obligatorily. The results of Experiments 3 and 4 showed that although both minima-and non-minima-segmented shapes pop out among unsegmented shapes, the search for minimasegmented shapes is significantly slower. Together, these results demonstrate that the visual system segments shapes into parts, using negative minima of curvature, and that it does so rapidly in early stages of visual processing.

Perception & Psychophysics, 1999

Visual completion is a ubiquitous phenomenon: Human vision often constructs contours and surfaces in regions that have no sharp gradients in any image property. When does human vision interpolate a contour between a given pair of luminance-defined edges? Two different answers have been proposed: relatability and minimizinginflections. Westate and prove a proposition that links these two proposals by showing that, under appropriate conditions, relatability is mathematically equivalent to the existence of a smooth curve with no inflection points that interpolates between the two edges. The proposition thus provides a set of necessary and sufficient conditions for two edges to be relatable. On the basis ofthese conditions, we suggest a way to extend the definition of relatability (1) to include the role of genericity, and (2) to extend the current all-or-none character of relatability to a graded measure that can track the gradedness in psychophysical data.

Perception & Psychophysics, 1999

Many researchers have proposed that, for the purpose of recognition, human vision parses shapes into component parts. Precisely how is not yet known. The minima rule for silhouettes defines boundary points at which to parse, but does not tell how to use these points to cut silhouettes, and therefore does not tell what the parts are. In this paper, we propose the short-cut rule, which states that, other things being equal, human vision prefers to use the shortest possible cuts to parse silhouettes. We motivate this rule, and the well-known Petters rule for modal completion, by the principle of transversality. We present five psychophysical experiments that test the short-cut rule, show that it successfully predicts part cuts which connect boundary points given by the minima rule, and show that it can also create new boundary points.

Proceedings of the National Academy of Sciences, 2006

Biol 38:205-287], because of their potential to provide a compact, but meaningful, shape representation, suitable for both neural modeling and computational applications. But effective computation of the shape skeleton remains a notorious unsolved problem; existing approaches are extremely sensitive to noise and give counterintuitive results with simple shapes. In conventional approaches, the skeleton is defined by a geometric construction and computed by a deterministic procedure. We introduce a Bayesian probabilistic approach, in which a shape is assumed to have ''grown'' from a skeleton by a stochastic generative process. Bayesian estimation is used to identify the skeleton most likely to have produced the shape, i.e., that best ''explains'' it, called the maximum a posteriori skeleton. Even with natural shapes with substantial contour noise, this approach provides a robust skeletal representation whose branches correspond to the natural parts of the shape.

Psychological Science, 1998

The perception of transparency is a remarkable feat of human vision: A single stimulation at the retina is interpreted as arising from two (or more) distinct surfaces, separated in depth, in the same visual direction. This feat is intriguing because physical transparency is neither necessary nor sufficient for phenomenal transparency. Many conditions for phenomenal transparency have been studied, including luminance, chromaticity, stereo depth, apparent motion, and structure from motion. Figural conditions have also been studied, primarily by Gestalt psychologists, resulting in descriptive laws. Here we extend, and make precise, these laws using the genericity principle and the minima rule for part boundaries. We report experiments that support the psychological plausibility of these refinements. The results suggest that the formation of visual objects and their parts is an early process in human vision that can precede the representation of transparency.

Perception & Psychophysics, 2003

Two experiments are reported in which we examined the ability of observers to identify landmarks on surfaces from different vantage points. In Experiment 1, observers were asked to mark the local maxima and minima of surface depth, whereas in Experiment 2, they were asked to mark the ridges and valleys on a surface. In both experiments, the marked locations were consistent across different observers and remained reliably stable over different viewing directions. These findings indicate that randomly generated smooth surface patches contain perceptually salient landmarks that have a high degree of viewpoint invariance. Implications of these findings are considered for the recognition of smooth surface patches and for the depiction of such surfaces in line drawings.

Perception, 2001

Journal of Archaeological Method and Theory, 2014

The Bronze Age/Iron Age transition in Prehistoric Europe represents a perfect case study to test different and competing hypotheses of social dynamics and economic change in small-scale societies. The paper discusses the possibilities of modeling what could have happened in Europe between 1800 and 800 BC, in terms of spatiotemporal dynamics. The paper presents some theoretical aspects of the dynamic study of expansive phenomena and gives an overview of a computer model programmed to explain the way new burial forms expanded in Europe. The main idea is comparing classic demic diffusion models (spread of population), cultural transmission models (spread of ideas), and technological innovation diffusion model (spread of goods). We will present the fundamentals of a preliminary study towards the computational simulation of such hypothetical social mechanisms, using a dataset composed of more than 1,500 georeferenced and radiocarbon dated archaeological contexts of a period between the Early Bronze Age and the first Iron Age (1800-800 BC) from an area including the North-East of Iberian

Computer-Aided Design and Applications, 2007

3D mesh segmentation is a fundamental process for Digital Shape Reconstruction in a variety of applications including Reverse Engineering, Medical Imaging, etc. It is used to provide a high level representation of the raw 3D data which is required for CAD, CAM and CAE. In this paper, we present an exhaustive overview of 3D mesh segmentation methodologies examining their suitability for CAD models. In particular, a classification of the various methods is given based on their corresponding underlying fundamental methodology concept as well as on the distinct criteria and features used in the segmentation process.

The Quarterly Journal of Experimental Psychology, 2009

Musicae Scientiae, 2014

Nineteen Temporal Semiotic Units (TSU) have been defined by musicians and researchers from the Laboratoire Musique et Informatique de Marseille (MIM) as categories of musical segments that acquire meaning through their temporal organization. The model of Parameterized Time Motifs (PTMs) provides a graphic modelling of TSUs, which is based on temporal functions and allows for their synthesis. The aim of the present study is to validate the model of PTMs through categorization. To do this, musician and non-musician participants were made to listen and group together three kinds of musical segments: TSUs taken from existing musical works, musical segments played on the piano that correspond to TSUs, and TSUs synthesized from the PTM model. Results have shown that each musical segment, whether synthesized from the PTM model or not, was more frequently paired with other segments of the same TSU category than with other TSU segments. No strong effect of musical expertise was observed. Finally, some TSU categories seemed to be more easily identifiable than others. The significance of these results is discussed in regard to the categorization theory and the effect of implicit learning on perception and knowledge.

Psychonomic Bulletin & Review, 2003

Memory & Cognition, 2005

How do we think about the space of bodies? Several accounts of mental representations of bodies were addressed in body part verification tasks. An imagery account predicts shorter times to larger parts (e.g., back Ͻ hand). A part distinctiveness account predicts shorter times to more discontinuous parts (e.g., arm Ͻ chest). A part significance account predicts shorter times to parts that are perceptually distinct and functionally important (e.g., head Ͻ back). Because distinctiveness and significance are correlated, the latter two accounts are difficult to distinguish. Both name-body and body-body comparisons were investigated in four experiments. In all, larger parts were verified more slowly than smaller ones, eliminating the imagery/size account. Despite the correlation between distinctiveness and significance, the data suggest that when comparisons are perceptual (body-body), part distinctiveness is the best predictor, and when explicit or implicit naming is involved, part significance is the best predictor. Naming seems to activate the functional aspects of bodies.

Proceedings of the National Academy of Sciences of the United States of America, 2014

Our visual environment abounds with curved features. Thus, the goal of understanding visual processing should include the processing of curved features. Using functional magnetic resonance imaging in behaving monkeys, we demonstrated a network of cortical areas selective for the processing of curved features. This network includes three distinct hierarchically organized regions within the ventral visual pathway: a posterior curvature-biased patch (PCP) located in the near-foveal representation of dorsal V4, a middle curvature-biased patch (MCP) located on the ventral lip of the posterior superior temporal sulcus (STS) in area TEO, and an anterior curvature-biased patch (ACP) located just below the STS in anterior area TE. Our results further indicate that the processing of curvature becomes increasingly complex from PCP to ACP. The proximity of the curvature-processing network to the well-known face-processing network suggests a possible functional link between them.

Perception, 1994

It is remarkable how well the human visual system can cope with changing viewpoints when it comes to recognising shapes. The state of the art in machine vision is still quite remote from solving such tasks. Nevertheless, a surge in invariance-based research has led to the development of methods for solving recognition problems still considered hard until recently. A nonmathematical account explains the basic philosophy and trade-offs underlying this strand of research. The principles are explained for the relatively simple case of planar-object recognition under arbitrary viewpoints. Well-known Euclidean concepts form the basis of invariance in this case. Introducing constraints in addition to that of planarity may further simplify the invariants. On the other hand, there are problems for which no invariants exist.

Behavioral and Brain Sciences, 1993

Lecture Notes in Computer Science, 2006

In this paper, we investigate human visual perception and establish a body of ground truth data elicited from human visual studies. We aim to build on the formative work of Ren, Eakins and Briggs who produced an initial ground truth database. Human subjects were asked to draw and rank their perceptions of the parts of a series of figurative images. These rankings were then used to score the perceptions, identify the preferred human breakdowns and thus allow us to induce perceptual rules for human decomposition of figurative images. The results suggest that the human breakdowns follow well-known perceptual principles in particular the Gestalt laws.

British Journal of Psychology, 2015

Humans have a preference for curved over angular shapes, an effect noted by artists as well as scientists. It may be that people like smooth curves or that people dislike angles, or both. We investigated this phenomenon in four experiments. Using abstract shapes differing in type of contour (angular vs. curved) and complexity, Experiment 1 confirmed a preference for curvature not linked to perceived complexity. Experiment 2 tested whether the effect was modulated by distance. If angular shapes are associated with a threat, the effect may be stronger when they are presented within peripersonal space. This hypothesis was not supported. Experiment 3 tested whether preference for curves occurs when curved lines are compared to straight lines without angles. Sets of coloured lines (angular vs. curved vs. straight) were seen through a circular or square aperture. Curved lines were liked more than either angular or straight lines. Therefore, angles are not necessary to generate a preference for curved shapes. Finally, Experiment 4 used an implicit measure of preference, the manikin task, to measure approach/avoidance behaviour. Results did not confirm a pattern of avoidance for angularity but only a pattern of approach for curvature. Our experiments suggest that the threat association hypothesis cannot fully explain the curvature effect and that curved shapes are, per se, visually pleasant.

PLoS ONE, 2013

Observers make sense of scenes by parsing images on the retina into meaningful objects. This ability is retained for line drawings, demonstrating that critical information is concentrated at object boundaries. Information theoretic studies argue for further concentration at points of maximum curvature, or corners, on such boundaries [1-3] suggesting that the relative positions of such corners might be important in defining shape. In this study we use patterns subtly deformed from circular, by a sinusoidal modulation of radius, in order to measure threshold sensitivity to shape change. By examining the ability of observers to discriminate between patterns of different frequency and/or number of cycles of modulation in a 2x2 forced choice task we were able to show, psychophysically, that difference in a single cue, the periodicity of the corners (specifically the polar angle between two points of maximum curvature) was sufficient to allow discrimination of two patterns near their thresholds for detection. We conclude that patterns could be considered as labelled for this measure. These results suggest that a small number of such labels might be sufficient to identify an object.

Japanese Psychological Research, 2011

The discriminative sensitivities of 30 4-month-old and 30 8-month-old infants for concave and convex objects were measured using the preferential-looking method. Five cylinder-like objects with different magnitudes of concave or convex shaded surfaces and outline contours were presented to the infants in pairs. The results indicated that the 4-month-old infants could discriminate better between object convexities than between object concavities. In contrast, the 8-month-old infants were able to equally discriminate between object concavities and object convexities, and their sensitivity to both object concavity/convexity was much higher than that of the 4-month-old infants. This difference in the sensitivity to object concavity and convexity suggested that younger and older infants might have differential abilities for cue utilization for recovering object structures.

Perception, 2003

The use of apertures for studying the perception of form has a long history. The main aim of the early investigations was to determine the conditions under which aperture viewing would lead to a complete percept of the underlying form (eg Zo« llner 1862; Helmholtz 1867/1962). A number of factors were found to be important, including the velocity with which the object was displaced behind the aperture or the speed of displacement of the aperture over the object, and the nature of the eye tracking that occurred during the movement of the object behind the aperture or the aperture over the object (eg Helmholtz 1867(eg Helmholtz /1962; for a review and summary, see see also Rock et al 1987).

Aesthetic Plastic Surgery, 2011

Assessment of facial soft tissues could be implemented using only anatomical landmarks. These points are so significant in the medical context because are able to provide significant information about the human face morphology and dimensions. At present their detection and location is made by expert physicians using palpation. Even if this procedure normally provides reliable information, the reliability of the results is proportional to the expertise of the physician. Considering that at present many physicians are beginning to use 3D scanners that provide three-dimensional data of the human face, it is possible to implement a robust and repeatable methodology that supports the physician's diagnosis. To reach this goal it is necessary to implement a methodology based on geometrical codification of landmarks and which mathematically formalizes the physician's visual and palpation analyses of the real patient. m female ALE

Attention, Perception, & Psychophysics, 2013

When a new visual object appears, attention is directed toward it. However, some locations along the outline of the new object may receive more resources, perhaps as a consequence of their relative importance in describing its shape. Evidence suggests that corners receive enhanced processing, relative to the straight edges of an outline (corner enhancement effect). Using a technique similar to that in an original study in which observers had to respond to a probe presented near a contour (Cole et al. in Journal of Experimental Psychology: Human Perception and Performance 27:1356-1368, 2001), we confirmed this effect. When figure-ground relations were manipulated using shaded surfaces (Exps. 1 and 2) and stereograms (Exps. 3 and 4), two novel aspects of the phenomenon emerged: We found no difference between corners perceived as being convex or concave, and we found that the enhancement was stronger when the probe was perceived as being a feature of the surface that the corner belonged to. Therefore, the enhancement is not based on spatial aspects of the regions in the image, but critically depends on figure-ground stratification, supporting the link between the prioritization of corners and the representation of surface layout.

Perception, 2006

Perception, 2004

Psychonomic Bulletin & Review, 2013

Interest in convexity has a long history in vision science. For smooth contours in an image, it is possible to code regions of positive (convex) and negative (concave) curvature, and this provides useful information about solid shape. We review a large body of evidence on the role of this information in perception of shape and in attention. This includes evidence from behavioral, neurophysiological, imaging, and developmental studies. A review is necessary to analyze the evidence on how convexity affects (1) separation between figure and ground, (2) part structure, and (3) attention allocation. Despite some broad agreement on the importance of convexity in these areas, there is a lack of consensus on the interpretation of specific claims-for example, on the contribution of convexity to metric depth and on the automatic directing of attention to convexities or to concavities. The focus is on convexity and concavity along a 2-D contour, not convexity and concavity in 3-D, but the important link between the two is discussed. We conclude that there is good evidence for the role of convexity information in figure-ground organization and in parsing, but other, more specific claims are not (yet) well supported.

Environment & Behavior, 2003

Human activity takes place in space. To act effectively, people need mental representations of space. People’s mental representations of space differ from space as conceived of by physicists, geometers, and cartographers. Mental representations of space are constructions based on elements, the things in space, and the spatial relations among them relative to a reference frame. People act in different spaces depending on the task at hand. The spaces considered here are the space of the body, the space around the body, the space of navigation, and the space of graphics. Different elements and spatial relations are central for functioning in the different spaces, yielding different mental representations.

Machine Vision and Applications

Matching deformable objects using their shapes is an important problem in computer vision since shape is perhaps the most distinguishable characteristic of an object. The problem is difficult due to many factors such as intra-class variations, local deformations, articulations, viewpoint changes and missed and extraneous contour portions due to errors in shape extraction. While small local deformations has been handled in the literature by allowing some leeway in the matching of individual contour points via methods such as Chamfer distance and Hausdorff distance, handling more severe deformations and articulations has been done by applying local geometric corrections such as similarity or affine. However, determining which portions of the shape should be used for the geometric corrections is very hard, although some methods have been tried. In this paper, we address this problem by an efficient search for the group of contour segments to be clustered together for a geometric correction using Dynamic Programming by essentially searching for the segmentations of two shapes that lead to the best matching between them. At the same time, we allow portions of the contours to remain unmatched to handle missing and extraneous contour portions. Experiments indicate that our method outperforms other algorithms, especially when the shapes to be matched are more complex.

Child Development, 2009

Three experiments investigated changes from 15 to 30 months of age in children's (N = 114) mastery of relations between an object and an aperture, supporting surface, or form. When choosing between objects to insert into an aperture, older children selected objects of an appropriate size and shape, but younger children showed little selectivity. Further experiments probed the sources of younger children's difficulty by comparing children's performance placing a target object in a hole, on a 2-dimensional form, or atop another solid object. Together, the findings suggest that some factors limiting adults' object representations, including the difficulty of comparing the shapes of positive and negative spaces and of representing shapes in 3 dimensions, contribute to young children's errors in manipulating objects.

Psychonomic Bulletin & Review, 2016

Seeing-perception and vision-is implicitly the fundamental building block of the literature on rationality and cognition. Herbert Simon and Daniel Kahneman's arguments against the omniscience of economic agents-and the concept of bounded rationality-depend critically on a particular view of the nature of perception and vision. We propose that this framework of rationality merely replaces economic omniscience with perceptual omniscience. We show how the cognitive and social sciences feature a pervasive but problematic meta-assumption that is characterized by an Ball-seeing eye.^We raise concerns about this assumption and discuss different ways in which the all-seeing eye manifests itself in existing research on (bounded) rationality. We first consider the centrality of vision and perception in Simon's pioneering work. We then point to Kahneman's work-particularly his article BMaps of Bounded Rationality^-to illustrate the pervasiveness of an all-seeing view of perception, as manifested in the extensive use of visual examples and illusions. Similar assumptions about perception can be found across a large literature in the cognitive sciences. The central problem is the present emphasis on inverse optics-the objective nature of objects and environments, e.g., size, contrast, and color. This framework ignores the nature of the organism and perceiver. We argue instead that reality is constructed and expressed, and we discuss the species-specificity of perception, as well as perception as a user interface. We draw on vision science as well as the arts to develop an alternative understanding of rationality in the cognitive and social sciences. We conclude with a discussion of the implications of our arguments for the rationality and decision-making literature in cognitive psychology and behavioral economics, along with suggesting some ways forward.

Perception, 1996

How well do observers perceive the local shape of an object from its shaded image? We address this complex problem by rst deriving a potential representation of local solid shape, and by presenting the results of a simple psychophysical experiment. Our descriptor of local solid shape, called shape characteristic, provides a viewpoint independent continuum between hyperbolic saddle-shaped and elliptic egg-shaped points. We then study the ability of human observers to make categorical judgments of local solid shape. We investigated this question using a smooth croissant", a simple object made of two connected regions of elliptic and hyperbolic points. Observers decided whether the surface was locally elliptic or hyperbolic at various points on the object. The task was natural, and the observers could reliably partition the shaded image of the object into one elliptic and one hyperbolic region. The ability of observers to perform this partition shows that they can, at least implicitly, localize the parabolic curves on a surface. This ability to locate the parabolic curve can in turn beexploited for other purposes, in particular to segment an object into its parts.

The Quarterly Journal of Experimental Psychology, 2015

Current theories of object recognition in human vision make different predictions about whether the recognition of complex, multipart objects should be influenced by shape information about surface depth orientation and curvature derived from stereo disparity. We examined this issue in five experiments using a recognition memory paradigm in which observers ( N = 134) memorized and then discriminated sets of 3D novel objects at trained and untrained viewpoints under either mono or stereo viewing conditions. In order to explore the conditions under which stereo-defined shape information contributes to object recognition we systematically varied the difficulty of view generalization by increasing the angular disparity between trained and untrained views. In one series of experiments, objects were presented from either previously trained views or untrained views rotated (15°, 30°, or 60°) along the same plane. In separate experiments we examined whether view generalization effects inter...

Nature Neuroscience, 2001

Inferotemporal (IT) cortex is involved in visual shape representation and visual object recognition, based on evidence from single-cell recording 1-9 , functional imaging 10,11 and lesion studies 9,12. In comparison with earlier visual areas, cells in IT have larger receptive fields and show more abstract preferences for complex shape properties 2,4,6,9,13 , but exactly how this region represents shape remains controversial 3-5,13. Here we examined shape representation within IT in relation to figure-ground reversal, as well as other stimulus manipulations that served as control comparisons. The figure-ground assignment of a given visual display can dramatically alter the shape that human observers perceive (examples, top of Fig. 1). Adjacent figure and ground regions defined by a common contour are perceived as very different. Human observers typically recognize the figure later (for example, the face in the top row of Fig. 1), but not the ground (white shape in that row), even for judgments based on exactly the same shared contour 14-18. Moreover, they rate a mirror image of the figure as more similar 19 to the original figure-ground display than an image of the ground in isolation. This arises even though the ground probe shares exactly the same curved contour as in the originally exposed display, whereas the mirror image of the figure has a mirrorreversed contour. These phenomena also arise for shapes made by unfamiliar contours 15-20 (see below), not just for profiles of meaningful shapes. Such effects reveal the influence of one-sided edge assignment on visual shape perception in humans 15-20. Here we examined how the shape preferences of IT cells in the primate brain may relate to these psychological phenomena. Specifically, we tested how the preferences of individual IT cells for stimuli drawn from a population of pseudorandom twodimensional shapes would generalize across three different transformations: figure-ground reversal, reversal of contrast-polarity and mirror-image reflection about the vertical (Fig. 1a-h). All shapes were polygons with straight edges at the top, bottom and along one side, and with a pseudo-randomly curved contour on

Visual Cognition, 2001

In five experiments, we investigated the detection of symmetry (i.e., translation plus reflection) or repetition (i.e., translation alone) between two vertical jagged contours. The complexity of the two contours was manipulated, as was their figure-ground assignment; the two contours either belonged to a common object "inside" them, to two separate objects "outside" them, or to two separate objects each to the right of one contour. Replicating Baylis and Driver (1994), symmetry judgements were unaffected by contour complexity when made within a single shape, implying a parallel process operating efficiently across contour discontinuities. However, all the other conditions yielded substantially slower judgements as complexity increased, suggesting either effortful point-by-poin t comparisons, or a highly inefficient parallel process. In agreement with Baylis and Driver (1995a), symmetry perception was harder when figure-ground assignment turned convexities along one contour into concavities along the other contour; and likewise for repetition detection. However, even when convex parts matched between the two contours, judgements were still affected by complexity unless they belonged to a common object. This supports Baylis and Driver's (1993) proposal that effortless comparisons for the layout of multiple convex parts can only be made within single perceptual objects. From a very young age, people are extremely good at detecting visual symmetry, especially about a vertical axis (e.g., Baylis, 1998; Bornstein & Krinsky, 1985). Symmetry judgements often occur swiftly and without apparent effort.

Perception & Psychophysics, 1995

CAAD futures 1997, 1997

The first stage of an investigation into the quantification and computability of architectural aesthetics is reported. Issues considered include the function, sources and role of aesthetic analysis in architecture in the framework of a descriptive approach to architectural analysis and design. The main focus is on the applicability of the concept of figural goodness to architectural aesthetics and the derivation of a representation for architectural form suitable to this purpose.

Design Computing and Cognition ’06, 2006

The notion of affordances has been used to represent functionality and usability in several design areas. The paper considers its applicability to architecture and buildings. It discusses a distinction between the affordances of building elements and spaces, and a number of dimensions for the mapping of different aspects.

Nature Neuroscience, 2004

Object perception depends on shape processing in the ventral visual pathway, which in monkeys culminates in inferotemporal cortex (IT). Here we provide a description of fundamental quantitative principles governing neural selectivity for complex shape in IT. By measuring responses to large, parametric sets of two-dimensional (2D) silhouette shapes, we found that neurons in posterior IT (Brodmann's areas TEO and posterior TE) integrate information about multiple contour elements (straight and curved edge fragments of the type represented in lower-level areas) using both linear and nonlinear mechanisms. This results in complex, distributed response patterns that cannot be characterized solely in terms of example stimuli. We explained these response patterns with tuning functions in multidimensional shape space and accurately predicted neural responses to the widely varying shapes in our stimulus set. Integration of contour element information in earlier stages of IT represents an important step in the transformation from low-level shape signals to complex object representation.

Seeing and perceiving, 2012

Rules and similarity are at the heart of our understanding of human categorization. However, it is difficult to distinguish their role as both determinants of categorization are confounded in many real situations. Rules are based on a number of identical properties between objects but these correspondences also make objects appearing more similar. Here, we introduced a stimulus set where rules and similarity were unconfounded and we let participants generalize category examples towards new instances. We also introduced a method based on the frequency distribution of the formed partitions in the stimulus sets, which allowed us to verify the role of rules and similarity in categorization. Our evaluation favoured the rule-based account. The most preferred rules were the simplest ones and they consisted of recurrent visual properties (regularities) in the stimulus set. Additionally, we created different variants of the same stimulus set and tested the moderating influence of small chang...

ACM Transactions on Graphics, 2013

Figure 1: Structural co-hierarchical analysis of a set of velocipedes (bicycles, tricycles and four-cycles). The resulting co-hierarchy (center) is illustrated by a single sample shape from the set, where each node represents a part assembly. Two of the nodes (highlighted in blue and green) are expanded to show the insight gained by the analysis which relates parts with rather different geometries but similar functions.

Vision Research, 1994

Subjects scanned line drawings of polygons in order to count tbe number of corners. The positions their eyes fixated were studied as a function of the size of tbe angle and whether tbe apex of tbe angle was present or absent. The results showed that the eyes tended to land at a position near tbe centre of gravity of the corner configurations. The observed landing positions were coherent with the hypothesis that tbe centre of gravity was calculated witbin an attentional spotlight centred on tbe apex of the corners, and that the calculation was based not on tbe total luminance distribution, nor on tbe distribution of energy in a neuropbysiologically motivated curvature detector, but simply on tbe basis of a contrast detector.

Lecture Notes in Computer Science, 2008

Decomposing a shape into meaningful components plays a strong role in shape-related applications. In this paper, we combine properties of skeleton and boundary to implement a general shape decomposition approach. It is motivated by recent studies in visual human perception discussing the importance of certain shape boundary features as well as features of the shape area; it utilizes certain properties of the shape skeleton combined with boundary features to determine protrusion strength. Experiments yield results similar to those from human subjects on abstract shape data. Also, experiments of different data sets prove the robustness of the combined skeleton-boundary approach.

Computer Graphics Forum, 2013

We define the convexity rank of a set of points to be the portion of mutually visible pairs of points out of the total number of pairs. Based on this definition of weak convexity, we introduce a spectral method that decomposes a given shape into weakly convex regions. The decomposition is applied without explicitly measuring the convexity rank. The method merely amounts to a spectral clustering of a matrix representing the all-pairs line of sight. Our method can be directly applied on an oriented point cloud and does not require any topological information, nor explicit concavity or convexity measures. We demonstrate the efficiency of our algorithm on a large number of examples and compare them qualitatively with competitive approaches.

Behavioral and Brain Sciences, 2010

Behavioral scientists routinely publish broad claims about human psychology and behavior in the world's top journals based on samples drawn entirely from Western, Educated, Industrialized, Rich, and Democratic (WEIRD) societies. Researchers-often implicitly-assume that either there is little variation across human populations, or that these "standard subjects" are as representative of the species as any other population. Are these assumptions justified? Here, our review of the comparative database from across the behavioral sciences suggests both that there is substantial variability in experimental results across populations and that WEIRD subjects are particularly unusual compared with the rest of the species-frequent outliers. The domains reviewed include visual perception, fairness, cooperation, spatial reasoning, categorization and inferential induction, moral reasoning, reasoning styles, self-concepts and related motivations, and the heritability of IQ. The findings suggest that members of WEIRD societies, including young children, are among the least representative populations one could find for generalizing about humans. Many of these findings involve domains that are associated with fundamental aspects of psychology, motivation, and behavior-hence, there are no obvious a priori grounds for claiming that a particular behavioral phenomenon is universal based on sampling from a single subpopulation. Overall, these empirical patterns suggests that we need to be less cavalier in addressing questions of human nature on the basis of data drawn from this particularly thin, and rather unusual, slice of humanity. We close by proposing ways to structurally reorganize the behavioral sciences to best tackle these challenges.

2021

Individual abilities in face recognition (good versus bad recognizers) were explored by means of event-related potentials (ERPs). The adaptation response profile of the N170 component to whole faces, eyes and mouths was used in order to highlight the crucial role of individual abilities in identity repetition processes for unfamiliar faces. The main point of this study is to underline the importance of characterizing the performance (bad or good) of the participants and to show that behaviorally selected groups might reveal neural differences. Good recognizers showed selective right hemisphere N170 repetition effects for whole faces and not for features. On the contrary, bad recognizers showed a general repetition effect not specifically related to faces and more pronounced processing for features. These findings suggest a different contribution of holistic and featural analysis in bad and good performers. In conclusion, we propose that the N170 might be used as a tool to tease apar...

Physics of Particles and Nuclei, 2010

We review the recent progress in the identification of graphical objects by the so called geometric correlation methods. We describe historical premises of these methods, and review the principles for design ing the features utilized in constructing the recognition methods. The identification methods themselves are also considered, and the quality of their performance is analyzed. Statistical methods for setting the classifi cation tolerance and the factors affecting the identification's precision are considered as well.

Symmetry, 2016

Primate vision is an active process that constructs a stable internal representation of the 3D world based on 2D sensory inputs that are inherently unstable due to incessant eye movements. We present here a mathematical framework for processing visual information for a biologically-mediated active vision stereo system with asymmetric conformal cameras. This model utilizes the geometric analysis on the Riemann sphere developed in the group-theoretic framework of the conformal camera, thus far only applicable in modeling monocular vision. The asymmetric conformal camera model constructed here includes the fovea's asymmetric displacement on the retina and the eye's natural crystalline lens tilt and decentration, as observed in ophthalmological diagnostics. We extend the group-theoretic framework underlying the conformal camera to the stereo system with asymmetric conformal cameras. Our numerical simulation shows that the theoretical horopter curves in this stereo system are conics that well approximate the empirical longitudinal horopters of the primate vision system.

Visual Cognition, 1998

This paper presents MetriCat, a model of the human capacity to recognize objects both as members of a general class (e.g., "chair") and as specific instances ("my office chair"), and of the role of visual attention in this capacity. MetriCat represents the attributes of an object's parts and their relations in a nonlinear fashion that provides a natural basis for recognition at both the class and instance levels (Stankiewicz & Hummel, 1996). Like previous structural description models (e.g., Hummel & Biederman, 1992), MetriCat represents part attributes and relations independently, dynamically binding them into structural descriptions. The resulting representation suggests two roles for visual attention in shape recognition: attention for binding and attention for signal-to-noise control. MetriCat implements both functions as special cases of a single mechanism for controlling the synchrony relations among units representing separate object parts. The model accounts for the time course of class-and instance-level classification, and makes several predictions about the relationships between attention, time, and levels of classification.

Behavioral and Brain Sciences, 1993

Fundamental to spatial knowledge in all species are the representations underlying object recognition, object search, and navigation through space. But what sets humans apart from other species is our ability to express spatial experience through language. This target article explores the language of objects and places, asking what geometric properties are preserved in the representations underlying object nouns and spatial prepositions in English. Evidence from these two aspects of language suggests there are significant differences in the geometric richness with which objects and places are encoded. When an object is named (i.e., with count nouns), detailed geometric properties-principally the object's shape (axes, solid and hollow volumes, surfaces, and parts)-are represented. In contrast, when an object plays the role of either "figure" (located object) or "ground" (reference object) in a locational expression, only very coarse geometric object properties are represented, primarily the main axes. In addition, the spatial functions encoded by spatial prepositions tend to be nonmetric and relatively coarse, for example, "containment," "contact," "relative distance," and "relative direction." These properties are representative of other languages as well. The striking differences in the way language encodes objects versus places lead us to suggest two explanations: First, there is a tendency for languages to level out geometric detail from both object and place representations. Second, a nonlinguistic disparity between the representations of "what" and "where" underlies how language represents objects and places. The language of objects and places converges with and enriches our understanding of corresponding spatial representations.

Skeletonization, 2017

Visual Cognition, 2010

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001, 2001

An improved method for deformable shape-based image segmentation is described. Image regions are merged together and/or split apart, based on their agreement with an a priori distribution on the global deformation parameters for a shape template. Perceptually-motivated criteria are used to determine where/how to split regions, based on the local shape properties of the region group's bounding contour. A globally consistent interpretation is determined in part by the minimum description length principle. Experiments show that model-guided split and merge yields a significant improvement in segmention over a method that uses merging alone.

Perception & Psychophysics, 1999

Wereport the results offour experiments that were conducted to examine both the representations that provide candidate entities available for object-based attentional selection and the influence of bottomup factors (i.e., geometric and surface characteristics of objects) and top-down factors (i.e. context and expectancies) on the selection process, Subjects performed the same task in each of the experiments. They were asked to determine whether two target properties, a bent end and an open end of a wrench, appeared in a brief display of two wrenches. In each experiment, the target properties could occur on a single wrench or one property could occur on each of two wrenches. The question of central interest was whether a same-object effect (faster and/or more accurate performance when the target properties appeared on one vs. two wrenches) would be observed in different experimental conditions, Several interesting results were obtained. First, depending on the geometric (i.e., concave discontinuities on object contours) and surface characteristics (i.e., homogeneous regions of color and texture) of the stimuli, attention was preferentially directed to one of three representational levels, as indicated by the presence or absence of the same-object effect. Second, although geometric and surface characteristics defmed the candidate objects available for attentional selection, top-down factors were quite influential in determining which representational level would be selected, Third, the results suggest that uniform connectedness plays an important role in defining the entities available for attention selection. These results are discussed in terms of the marmer in which attention selects objects in the visual environment.

Design, User Experience, and Usability: Users and Interactions, 2015

Multi-modal interfaces that provide haptic access to statistical line graphs combined with verbal assistance are proposed as an effective tool to fulfill the needs of visually impaired people. Graphs do not only present data, they also provide and elicit the extraction of second order entities (such as maxima or trends), which are closely linked to shape properties of the graphs. In an experimental study, we investigated collaborative joint activities between haptic explorers of graphs and verbal assistants who helped haptic explorers to conceptualize local and non-local second-order concepts. The assistants have not only to decide what to say but in particular when to say it. Based on the empirical data of this experiment, we describe in the present paper the design of a feature set for describing patterns of haptic exploration, which is able to characterize the need for verbal assistance during the course of haptic exploration. We employed a (supervised) classification algorithm, namely the J48 decision tree. The constructed features within the range from basic (low-level) user-action features to complex (high-level) conceptual were categorized into four feature sets. All feature set combinations achieved high accuracy level. The best results in terms of sensitivity and specificity were achieved by adding the low-level graphical features.

2019

Object memories activated by borders serve as priors for figure assignment: figures are more likely to be perceived on the side of a border where a well-known object is sketched. Do object memories also affect the appearance of object borders? Memories represent past experience with objects; memories of well-known objects include many with sharp borders because they are often fixated. We investigated whether object memories affect appearance by testing whether blurry borders appear sharper when they are contours of well-known objects versus matched novel objects. Participants viewed blurry versions of one familiar and one novel stimulus simultaneously for 180ms; then made comparative (Exp. 1) or equality judgments regarding perceived blur (Exps. 2-4). For equivalent levels of blur, the borders of well-known objects appeared sharper than those of novel objects. These results extend evidence for the influence of past experience to object appearance, consistent with dynamic interactive...

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

We study the problem of compositional zero-shot learning for object-attribute recognition. Prior works use visual features extracted with a backbone network, pre-trained for object classification and thus do not capture the subtly distinct features associated with attributes. To overcome this challenge, these studies employ supervision from the linguistic space, and use pre-trained word embeddings to better separate and compose attribute-object pairs for recognition. Analogous to linguistic embedding space, which already has unique and agnostic embeddings for object and attribute, we shift the focus back to the visual space and propose a novel architecture that can disentangle attribute and object features in the visual space. We use visual decomposed features to hallucinate embeddings that are representative for the seen and novel compositions to better regularize the learning of our model. Extensive experiments show that our method outperforms existing work with significant margin on three datasets: MIT-States, UT-Zappos, and a new benchmark created based on VAW. The code, models, and dataset splits are publicly available at https: //github.com/nirat1606/OADis.

Environment and Planning B: Planning and Design, 2012

Embedding parts is a key problem in computing when dealing with continuous matter such as shapes rather than discrete matter such as symbols. For computing part relations such as embedding, a technical framework that uses weighted shapes is introduced and implemented. In the proposed framework, for any given two-dimensional shape, the entire canvas is defined as a weighted shape and serves as a registration mark in detecting embedded parts. The approach treats shapes as perceived wholes rather than composed and eliminates the technical distinction between shape categories such as line, curve, or plane. The implementation is shown for two-dimensional shapes but is extendable to three dimensions. As demonstrated on a Seljuk geometric pattern, the framework allows for embedding multiple and various perceived wholes, thus exploring emerging shapes and shape relations to be used for analysis and synthesis in design.

Cognitive Science, 1989

In a recent paper, Chambers and Reisberg (1985) showed that people cannot reverse classical ambiguous

Attention, Perception, & Psychophysics, 2018

Regularities like symmetry (mirror reflection) and repetition (translation) play an important role in both visual and haptic (active touch) shape perception. Altering figure-ground factors to change what is perceived as an object influences regularity detection. For vision, symmetry is usually easier to detect within one object, whereas repetition is easier to detect across two objects. For haptics, we have not found this interaction between regularity type and objectness (

Signal, Image and Video Processing, 2017

Liver is considered as a tissue with large shape variations. Representation of such a complex shape cannot be accurately performed using conventional SSM algorithms. We propose to decompose a human liver into its anatomical parts and use them as a guide when considering point correspondences. To cope with shape complexity, a modified coherent point drift (CPD) algorithm is proposed too. The modified CPD algorithm assigns fuzzy correspondences to points and follows a simulated annealing approach to convert fuzzy correspondences into binary ones. Our modification includes automatic parameter settings which results in robustness of the algorithm. The proposed algorithm was compared to the thin plate spline-robust point matching (TPS-RPM) and minimum description length (MDL) techniques. Our method is twice faster than the MDL algorithm. Compared to the TPS-RPM algorithm, our method improved mean Specificity of the right lobe by 0.09 and Compactness of the model by two less modes.

Multimedia Tools and Applications, 2017

Understanding the multimodal rendering of 3D shapes is becoming an important research topic as multimedia and virtual reality technologies are rapidly advancing. This study is aimed to investigate human perceptibility on the curvature and texture changes of 3D virtual surfaces across modalities, vision and touch. Our interest is to obtain perception data that can be used for 3D watermarking or data compression under a virtual reality environment providing multimodal interactions. For this study, we designed two psychophysical experiments to estimate curvature discrimination and texture detection thresholds on curvature surfaces over three conditions: vision only, touch only, and both vision and touch. The results show that touch is dominant at both discriminating curvature surfaces and detecting surface texture changes on a curved surface. In addition, the sensitivity of the both senses to detect texture changes linearly increases as a curvature value increases. Finally, the vision and touch senses compensate each other when both modalities are available at the same time. The thresholds from the present study can potentially be used as the upper limit for selecting watermark strengths or compression in order to ensure imperceptibility in a 3D visuohaptic multimedia systems.

2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06)

A combination of techniques that is becoming increasingly popular is the construction of part-based object representations using the outputs of interest-point detectors. Our contributions in this paper are twofold: first, we propose a primal-sketch-based set of image tokens that are used for object representation and detection. Second, top-down information is introduced based on an efficient method for the evaluation of the likelihood of hypothesized part locations. This allows us to use graphical model techniques to complement bottom-up detection, by proposing and finding the parts of the object that were missed by the front-end feature detection stage. Detection results for four object categories validate the merits of this joint top-down and bottom-up approach.

Computer Vision – ECCV 2018, 2018

The world of human-object interactions is rich. While generally we sit on chairs and sofas, if need be we can even sit on TVs or top of shelves. In recent years, there has been progress in modeling actions and human-object interactions. However, most of these approaches require lots of data. It is not clear if the learned representations of actions are generalizable to new categories. In this paper, we explore the problem of zero-shot learning of human-object interactions. Given limited verb-noun interactions in training data, we want to learn a model than can work even on unseen combinations. To deal with this problem, In this paper, we propose a novel method using external knowledge graph and graph convolutional networks which learns how to compose classifiers for verbnoun pairs. We also provide benchmarks on several dataset for zero-shot learning including both image and video. We hope our method, dataset and baselines will facilitate future research in this direction.

Perception & Psychophysics, 2001

Using the visual search method, we show that stimuli that contain curvature discontinuities (i.e., points where the second derivative along an image contour is not defined) are easily found among stimuli containing only smooth changes in curvature. Curved stimuli that lack curvature discontinuities, however, are difficult to find among distractors that have them. These results suggest that the visual system detects and analyzes abrupt changes in curvature in the image quickly to extract vital information about the 3-D structure of the visual environment.

Cognition

Previous research has shown that humans are able to acquire statistical regularities among shape parts that form various spatial configurations, via exposure to these configurations without any task or feedback. The present study extends this approach of visual statistical learning to examine whether prior knowledge of parts, acquired in a separate learning context, facilitates acquisition of multi-layer hierarchical representations of objects. After participants had learned to encode a shape-pair as a chunk into memory, they viewed cluttered scenes containing multiple shape chunks. One of the larger configurations was constructed by combining the learned shape-pair with an unfamiliar, complementary shape-pair. Although the complementary shape-pair had never been presented separately during learning, it was remembered better than other shape pairs that were parts of larger configurations. The greater perceived familiarity of the complementary shape-pair depended on the encoding strength of the previously learned shape-pair. This "parts-beget-parts" effect suggests that statistical learning, in combination with prior knowledge, can represent objects as a coherent whole and also as a spatial configuration of parts by bootstrapping multi-layer hierarchical structures.

Attention, Perception, & Psychophysics, 2010

Research indicates that object perception involves the decomposition of images into parts. A critical principle that governs part decomposition by adults is the shortcut rule, which states that, all else being equal, the visual system parses objects using the shortest possible cuts. We examined whether 6.5-month-olds' parsing of images also follows the shortcut rule. Infants in the experimental conditions were habituated to cross shapes and then tested for their preference between segregated patterns produced using long cuts versus short cuts. Infants in the control conditions were directly tested with the segregated patterns. Infants in the experimental conditions exhibited a greater novelty preference for the long-cut over the shortcut patterns than did those in the control conditions, thereby indicating that they are more likely to segregate cross shapes using short cuts rather than long cuts. This sensitivity to the shortcut rule was evident when two alternative parameters, part area and protrusion, were controlled in Experiments 1 and 2, respectively. Thus, a critical principle that governs part segregation in adulthood is operational by 6.5 months of age.

Revue de Synthèse, 1990

Le dualisme symbolique/physique 1. Pour plus de precisions concernant les references placees entre parentheses, dans cet article, se reporter ala Bibliographie, p. 180.

Perception, 2008

Introduction Line drawings of everyday objects are generally very easy to identify, perhaps even as easy as the original objects themselves, which shows that the most essential information about object identity is preserved (or even enhanced) in line drawings (Biederman 1987; Biederman and Ju 1988; Koenderink and van Doorn 1982). Moreover, cave art, cartoons, and comic books go one step further in demonstrating that the lines do not have to capture the edges defined by luminance or colour contrast very accurately to reveal a particular prehistoric animal or political figure. Cave art has attracted the attention of perception researchers mainly for this reason (Gerr 1982; Kennedy and Silver 1974; Wade 2006; Wade and Melcher 2006). Combined image-processing and psychophysical research has shown that line drawings are indeed much more informative than the edges extracted from grey-level photographs (Sanocki et al 1998). The best artists can extract the essence of an object's shape or a person's facial characteristics from the rich sources of visual information in natural viewing conditions or in highquality photographs and express these in a few lines on a flat surface like a piece of paper. The fact that this process works so well suggests that this kind of information (ie a few high-contrast, curved lines) triggers the kind of machinery that the visual system has developed (phylogenetically and/or ontogenetically) to extract information about object shape and identity from the visual stimulation that is available in natural viewing conditions (Braje et al 1995; Koenderink et al 1996). It is for all these reasons that line drawings of everyday objects have become a widely used type of stimulus in experimental research on object identification and picture naming. Snodgrass and Vanderwart (1980) have given a strong additional impetus to this kind of research by providing a set of 260 line drawings with published norms for name agreement, familiarity, complexity, etc, which has been very instrumental in comparing results on picture naming, object identification, priming, etc across studies.

Lecture Notes in Computer Science, 2007

In this paper, we propose an original framework for three dimensional face representation and similarity matching. Basic traits of a face are encoded by extracting convex and concave regions from the surface of a face model. A compact graph representation is then constructed from these regions through an original modeling technique capable to quantitatively measure spatial relationships between regions in a three dimensional space and to encode this information in an attributed relational graph. In this way, the structural similarity between two face models is evaluated by matching their corresponding graphs. Experimental results on a 3D face database show that the proposed solution attains high retrieval accuracy and is reasonably robust to facial expression and pose changes.

Behavioral and Brain Sciences, 1993

Fundamental to spatial knowledge in all species are the representations underlying object recognition, object search, and navigation through space. But what sets humans apart from other species is our ability to express spatial experience through language. This target article explores the language of objects and places, asking what geometric properties are preserved in the representations underlying object nouns and spatial prepositions in English. Evidence from these two aspects of language suggests there are significant differences in the geometric richness with which objects and places are encoded. When an object is named (i.e., with count nouns), detailed geometric properties -principally the object's shape (axes, solid and hollow volumes, surfaces, and parts) -are represented. In contrast, when an object plays the role of either "figure" (located object) or "ground" (reference object) in a locational expression, only very coarse geometric object properties are represented, primarily the main axes. In addition, the spatial functions encoded by spatial prepositions tend to be nonmetric and relatively coarse, for example, "containment," "contact," "relative distance," and "relative direction." These properties are representative of other languages as well. The striking differences in the way language encodes objects versus places lead us to suggest two explanations: First, there is a tendency for languages to level out geometric detail from both object and place representations. Second, a nonlinguistic disparity between the representations of "what" and "where" underlies how language represents objects and places. The language of objects and places converges with and enriches our understanding of corresponding spatial representations.

Representations, Analysis and Recognition of Shape and Motion from Imaging Data, 2017

Here, we intend to propose a 2D contour descriptor that we call Generalized Curvature Scale Space (GCSS) based on the isocurvature levels, and the curvature scale space (CSS) descriptor. We start by computing the curvature in different scales and extract the points which have the same curvature values as the maximums in each scale. Each CSS image is represented by a set of key points. The Dynamic Time Warping (DTW) similarity measure is used. We reach a significant rate in image recognition using two data sets (HMM GPD and MPEG7 CE Shape-1 Part-B set).

Representations, Analysis and Recognition of Shape and Motion from Imaging Data, 2019

Here, we intend to propose a discrete normalization of the Generalized Curvature Scale Space (GCSS). The GCSS is an Euclidean invariant planar contour descriptor. It consists on the convolution of the contour by Gaussian functions with different scales. The points having the same curvature values as the selected extremums are the considered key points. This representation implies different number of descriptors from a shape to another. Thus, a step of redistribution of the key points is requested. Therefore, a discrete normalization approach is proceeded. The descriptor is composed by the curvature variation of the key points at the smoothed curve. Several datasets were used to carry on the experiments and to verify the accuracy, the stability and the robustness of the novel description. The Dynamic Time Warping distance is the similarity metric used. Experimental results show that considerable rates of image retrieval are reached comparing to the state of the art.

Philosophy and Phenomenological Research, 2011

Negative facts get a bad press. One reason for this is that it is not clear what negative facts are. We provide a theory of negative facts on which they are no stranger than positive atomic facts. We show that none of the usual arguments hold water against this account. Negative facts exist in the usual sense of existence and conform to an acceptable Eleatic principle. Furthermore, there are good reasons to want them around, including their roles in causation, chance-making and truth-making, and in constituting holes and edges. . There also exist arguments against facts in general such as the infamous slingshot argument, which purports to show that there is only one fact (see, e.g., Neale ). We think that the slingshot is over-rated, as do many others (see Restall ). . Parsons ( , -) expresses doubts about the coherence of describing entities as 'positive' or 'negative'.