Veridicality by simplicity

A principal challenge for both human and machine vision systems is to integrate and organize the diversity of cues received from the environment into the coherent global representations required to make good decisions and take effective actions. This conference brings together an interdisciplinary roster of leading researchers in both biological and computer vision to report and discuss the latest research on this process of perceptual organization.

We perceive structure through a process of perceptual organization. Here we report a new perceptual organization phenomenon-the facilitation of visual grouping by global curvature. Observers viewed patterns that they perceived as organized into collections of curves. The patterns were perceptually ambiguous such that the perceived orientation of the patterns varied from trial to trial. When patterns were sufficiently dense and proximity was equated for the predominant perceptual alternatives, observers tended to perceive the organization with the greatest curvature. This effect is tantamount to visual grouping by maximal curvature and thus demonstrates an unprecedented effect of global structure on perceptual organization. We account for this result with a model that predicts the perceived organization of a pattern as function of its nonaccidentality, which we define as the probability that it could have occurred by chance. Our findings demonstrate a novel relationship between the geometry of a pattern and the visual salience of global structure.

The processes of perceptual organization and social categorization share the goal of simplifying and structuring the world, though they have traditionally been studied within separate branches of psychology. Drawing on embodied cognition, this perspective suggests that both processes stem from bodily experiences. Early sensorimotor experiences, such as being held in proximity or rocked in synchrony, may have shaped the principles of perceptual organization and provide the foundations for social categorization. To integrate these domains, a triadic embodied model is proposed, linking bodily experience, perceptual organization, and social categorization as dynamically interacting vertices. The model allows bidirectional influences, whereby each element can prime and reinforce the others, and may also involve activation of metaphorical concepts. It further explains how cultural context can bias these interconnections. The model situates social categorization along a continuum of representational depth, from shallow symbolic tagging to deep embodied simulation. While prior work has demonstrated dyadic links between bodily experience, perceptual organization, and social categorization, the present article explicitly integrates these domains in a unified model, offering new directions for understanding how people structure their physical and social world.

Some resularities are more salient to the visual system than others. P. A. van der Helm and E. L. J. Leeuwenberg (1996) have proposed a new approach that quantifies the goodness of a pattern's regularity as the number of holographic identities constituting the regulartiy, relative to the total amount of information needed to describe the pattern. This holographic approach to goodness was compared with previous approaches and was presented in relation to metatheoretic al issues. Tnese 3 aspects are discussed further here. First, the theory is shown to contain implausible assumptions and unfortunate gaps with respect to the required processing. Second, Van der Helm and Lseuwenberg's critique on preceding theories is refuted. Third, some metatheorerical issues need to be qualified or at least clarified. Together, these concerns suggest that a better approach to goodness might result from a synthesis of the most useful aspects of diverse theories of goodness. In 1996, Van der Helm and Leeuwenberg presented a new theory of goodness that helped researchers understand the perceptual salience of regularities such as repetition and mirror symmetry in terms of properties of their mathematical representation. They argued that preceding theories of goodness failed either because they were process theories with contradictory ad hoc assumptions or because they were representation theories taking a transformational approach to goodness. In contrast, their theory is representation based and nontransformational. First, I summarize Van der Helm and Leeuwenberg's (1996) theory and discuss some problematic assumptions and unfortunate gaps. Then, I demonstrate that their critique of preceding theories such as the transformational approach and the bootstrap model is not completely justified. Finally, I raise some concerns about the metatheoretical framework within which their theory has been presented. The major message is that a synthesis of the most fruitful aspects of the new representation theory with those of preceding process theories might lead toward a better approach to goodness.

In this work, we discussed and counter-commented van der Helm’s comments on our previous paper (Pinna and Conti, Brain Sci., 2019, 9, 149), where we demonstrated unique and relevant visual properties imparted by contrast polarity in eliciting amodal completion. The main question we addressed was: “What is the role of shape formation and perceptual organization in inducing amodal completion?” To answer this question, novel stimuli were studied through Gestalt experimental phenomenology. The results demonstrated the domination of the contrast polarity against good continuation, T-junctions, and regularity. Moreover, the limiting conditions explored revealed a new kind of junction next to the T- and Y-junctions, respectively responsible for amodal completion and tessellation. We called them I-junctions. The results were theoretically discussed in relation to the previous approaches and in the light of the phenomenal salience imparted by contrast polarity. In counter-commenting van der Helm’s comments we went into detail of his critiques and rejected all of them point-by-point. We proceeded by summarizing hypotheses and discussion of the previous work, then commenting on each critique through old and new phenomena and clarifying the meaning of our previous conclusions.

Perceptual organization concerns the organization of people’s perception into meaningful, structured entities—often wholes containing parts, with specific interrelationships among the parts as well as specific relationships between the parts and the wholes. It embraces a wide variety of distinct processes, which often occur together in everyday life and lead to interesting perceptual experiences by their interactions. The best-known perceptual organization processes are perceptual grouping and figure-ground organization, which appear to be determined by characteristic principles or laws, such as grouping by proximity, similarity, good continuation, area, convexity, symmetry, and so forth. Other important phenomena of perceptual organization include perceptual multi-stability (i.e., the fact that the same stimulus can give rise to multiple competing percepts that switch over time), visual illusions, the role of reference frames, and the role of context. Gestalt psychology has contributed deep insights into the way people’s perception of the world is organized, although it has also been criticized for its emphasis on subjective experiences and its sometimes vague and non-quantitative descriptions of the Gestalt principles. Because perceptual organization essentially concerns visual awareness, however, phenomenology is needed to describe what is experienced. This ultimately remains the basis for research on perceptual organization even when it is aimed at quantifying the principles of perceptual organization and identifying the neural correlates or mechanisms underlying perceptual organization.

Schizophrenia is associated with impaired visual information processing. The aim of this study was to investigate the relationship between anomalous perceptual experiences, positive and negative symptoms, perceptual organization, rapid categorization of natural images and magnocellular (M) and parvocellular (P) visual pathway functioning. Thirty-five unmedicated patients with schizophrenia and 20 matched healthy control volunteers participated. Anomalous perceptual experiences were assessed with the Bonn Scale for the Assessment Basic Symptoms (BSABS). General intellectual functions were evaluated with the revised version of the Wechsler Adult Intelligence Scale. The 1-9 version of the Continuous Performance Test (CPT) was used to investigate sustained attention. The following psychophysical tests were used: detection of Gabor patches with collinear and orthogonal flankers (perceptual organization), categorization of briefly presented natural scenes (rapid visual processing), low-contrast and frequency-doubling vernier threshold (M pathway functioning), isoluminant colour vernier threshold and high spatial frequency discrimination (P pathway functioning). The patients with schizophrenia were impaired on test of perceptual organization, rapid visual processing and M pathway functioning. There was a significant correlation between BSABS scores, negative symptoms, perceptual organization, rapid visual processing and M pathway functioning. Positive symptoms, IQ, CPT and P pathway measures did not correlate with these parameters. The best predictor of the BSABS score was the perceptual organization deficit. These results raise the possibility that multiple facets of visual information processing deficits can be explained by M pathway dysfunctions in schizophrenia, resulting in impaired attentional modulation of perceptual organization and of natural image categorization.

Part I : Perceptual Organization and Classification Processes

Perceptual organization in task performance

Perceptual organization remains one of the very few established theories on the human visual system. It underpinned many pre-deep seminal works on segmentation and detection, yet research has seen a rapid decline since the preferential shift to learning deep models. Of the limited attempts, most aimed at interpreting complex visual scenes using perceptual organizational rules. This has however been proven to be sub-optimal, since models were unable to effectively capture the visual complexity in real-world imagery. In this paper, we rejuvenate the study of perceptual organization, by advocating two positional changes: (i) we examine purposefully generated synthetic data, instead of complex real imagery, and (ii) we ask machines to synthesize novel perceptually-valid patterns, instead of explaining existing data. Our overall answer lies with the introduction of a novel visual challenge – the challenge of perceptual question answering (PQA). Upon observing example perceptual question-answer pairs, the goal for PQA is to solve similar questions by generating answers entirely from scratch (see Figure 1). Our first contribution is therefore the first dataset of perceptual question-answer pairs, each generated specifically for a particular Gestalt principle. We then borrow insights from human psychology to design an agent that casts perceptual organization as a self-attention problem, where a proposed grid-to-grid mapping network directly generates answer patterns from scratch. Experiments show our agent to outperform a selection of naive and strong baselines. A human study however indicates that ours uses astronomically more data to learn when compared to an average human, necessitating future research (with or without our dataset).

The review focuses on the evolution of views on the mechanisms of human colour perception in context of a gradual study of the spectral characteristics of light, the physiology and psychology of the visual mechanism, and the decoding of the functional features of retinal sensors and other neural groups throughout the visual pathway. The first part of the review presents historical information on the various stages of the development of views on the nature of colour vision not only from the perspective of natural science hypotheses provided by scientists, but also regarding the subjective opinions of representatives from the arts-artists and poets. The review briefly outlines concepts of ancient chromatism and medieval metaphysics of colour and continues with the first rational theories of colour perception, providing more detailed discussions on the views of Isaac Newton and Mikhail Lomonosov, Johann Goethe and John Turner, Gaspard Monge and Hermann von Helmholtz, Arthur Schopenhauer and Erwin Schrödinger, James Maxwell, and Ewald Hering. The second and third parts of the review will sequentially examine the progress in understanding the operation of colour-coding neurons, their receptive fields, colour discrimination thresholds, and other aspects in relation to such effects as colour constancy, the relationship between the form of objects and their perceived colour, and various colour illusions. Recent intriguing results, obtained using molecular genetic methods, shed new light on the evolutionary transformations of colour vision in animals and humans. The review will show an impact of the major studies of deviations in colour perception on practically significant lighting solutions, discussion of the individualities of colour perception in virtual and augmented reality systems, as well as in context of various holographic technologies. Furthermore, the matter of museum lighting optimization and the production of specialized LED devices will be considered. Finally, the review presents the link between the latest achievements in the field of psychophysiology of colour perception and the spectral characteristics of light sources familiar to specialists in colourimetry and lighting technology.

Objective To explore the characteristics of intelligence characteristics in preschool children with attention deficit hyperactivity disorder (ADHD). Methods From June 2007 to May 2010, 43 preschool children at the age of 4 to 6 with attention deficit hyperactivity disorder and 32 healthy preschool children at the same age were recruited in this study. Their intelligence was evaluated with Chinese-Wechsler Young Children Scale of Intelligence (C-WYCSI). Informed consent was obtained from all participates. Results The verbal intelligence quotient (VIQ), performance intelligence quotient (PIQ) and full intelligence quotient (FIQ) of preschool children with attention deficit hyperactivity disorder were at the normal range, but were significantly lower than those of control group (P<0.05). Children with attention deficit hyperactivity disorder had significantly lower scores of information, arithmetic, vocabulary, picture concepts, comprehension, AE, Ma and block design than those of control group (P<0.05). Twenty-two children with attention deficit hyperactivity disorder and 3 children in control group had discrete scores between verbal intelligence quotient and performance intelligence quotient (≥15 scores). The difference of children who had discrete scores between verbal intelligence quotient and performance intelligence quotient was significant (P<0.05). In factor analysis, the scores of verbal comprehension, perceptual organization an, visual stimuli comprehension, simultaneous processing, following processing and action plan of ADHD group were significantly lower than that of the controls (P<0.05). Conclusion Preschool children with attention deficit hyperactivity disorder have lower intelligence quotient (IQ) scores than that of normal children and their balance between verbal intelligence quotient and performance intelligence quotient was poor. Key words: attention deficit hyperactivity disorder(ADHD); intelligence; preschool; child

Every object acquires its shape from some kind of generative process, such as manufacture, biological growth, or self-organization, in response to external forces. Inferring such generative processes from an observed shape is computationally challenging because a given process can lead to radically different shapes, and similar shapes can result from different generative processes. Here, we suggest that in some cases, generative processes endow objects with distinctive statistical features that observers can use to classify objects according to what has been done to them. We found that from the very first trials in an eight-alternative forced-choice classification task, observers were extremely good at classifying unfamiliar objects by the transformations that had shaped them. Further experiments show that the shape features underlying this ability are distinct from Euclidean shape similarity and that observers can separate and voluntarily respond to both aspects of objects. Our findings suggest that perceptual organization processes allow us to identify salient statistical shape features that are diagnostic of generative processes. By so doing, we can classify objects we have never seen before according to the processes that shaped them.

The processing and perception of visual stimuli are greatly dependent on the context in which those stimuli are presented. While this is not always (and indeed is probably not supposed to be) evident when viewing real-world scenes, the abundance of contextual effects is revealed for example by the vast number of different visual illusions. Context can profoundly change the way the visual system processes single visual stimuli. In contrast masking, surround suppression, and crowding, distracting stimuli impair the observer's ability to detect, identify, and make other judgements about visual stimuli. These perceptual phenomena presumably reflect the operation of neural mechanisms that adjust the operating range of the visual system according to the prevailing conditions, or act to reduce responses to redundant stimulus features. This work first investigates the relationship between masking, which is caused by stimuli that are spatially superimposed on a target stimulus, and surround suppression, which is caused by spatially displaced stimuli. This was done using a temporal masking paradigm and measuring the time-courses of masking and surround suppression. The results reveal how the effects of these suppressive interactions are delicately dependent on stimulus timing and suggest that surround suppression acts extremely fast, having an effect on the early part of the neural response. Finally, contextual effects such as surround suppression and crowding are often considered to be automatic processes that occur over fixed areas in the visual field. The final part of this work shows how changing the spatial layout, and consequently the perceptual organization, of the stimuli can dramatically reduce surround suppression and crowding effects, even in conditions where the current models would predict strong effects. Surround suppression and crowding seem to operate only within perceptually homogeneous areas, while stimuli that are perceptually separate, the ones that stand out, are not affected.

The science and technology of image transmission and display has evolved primarily from the diciplines of computer science, electrical engineering and physics. Thus, it is only natural that techniques and attitudes which have developed are characterized by the style in which people in these areas approach the subject of images and their transmission, display, and processing. For example, one finds many important differences between the methods of television and those of photography. Moreover, an even greater contrast is found when comparing the methods of electrical engineering, physics, television and photography with those now understood to be employed by the human eye and the human visual system. Specifically, let us contrast the method by which television, photography and the human visual system, represent image information. In television (particularly digital imaging), image values are represented by signals analogous to quantities of light. They are called intensities. In photography, on the other hand, the representing quantities are concentrations of silver or dyes. Consequently, due to the natural exponentiating laws governing the interaction of light with these media, they are analogous to the logarithm of quantities of light. They are called densities. The human visual system, on the other hand, (while being generally logarithmically sensitive) moves a large step further away from representation by physical quantities of light and produces at very early stages in its processing highly modified versions of the patterns of light or their logarithms. The natural question then arrises, why should the human visual system try to do this; and since it does, what consequences are implied in terms of the television and photographic presentations normally employed?