Building Emerging Images with Tiled Orders and Posets

We introduce an algorithm based on posets and tiled orders to generate emerging images. Experimental results allow concluding that images obtained with these kinds of tools are easy to detect by human beings. It is worth pointing out that the emergence phenomenon is a Gestalt grouping law associated with AI open problems. For this reason, emerging images have arisen in the last few years as a tool in the context of the development of human interactive proofs.

Does perceptual grouping improve visuospatial working memory? Optimized processing or encoding bias

Visual working memory has been defined as a system of limited capacity that enables the maintenance and manipulation of visual information. However, some perceptual features like Gestalt grouping could improve visual working memory effectiveness. In two different experiments, we aimed to explore how the presence of elements grouped by color similarity affects the change detection performance of both, grouped and non-grouped items. We combined a change detection task with a retrocue paradigm in which a six item array had to be remembered. An always valid, variable-delay retrocue appeared in some trials during the retention interval, either after 100 ms (iconic-trace period) or 1400 ms (working memory period), signaling the location of the probe. The results indicated that similarity grouping biased the information entered into the visual working memory, improving change detection accuracy only for previously grouped probes, but hindering change detection for non-grouped probes in certain conditions (Exp. 1). However, this bottom-up automatic encoding bias was overridden when participants were explicitly instructed to ignore grouped items as they were irrelevant for the task (Exp. 2).

EXPRESS: Competition between perceptual grouping cues in an indirect objective task

The integration between Gestalt grouping cues has been a relatively unexplored issue in vision science. The present work introduces an objective indirect method based on the repetition discrimination task to determine the rules that govern the dominance dynamics of the competition between both intrinsic (Experiment 1: proximity vs. luminance similarity) and extrinsic grouping cues (Experiment 2: common region vs. connectedness) by means of objective measures of grouping (reaction times and accuracy). Prior to the main task, a novel objective equating task was introduced with the aim of equating the grouping strength of the cues for the visuomotor system. The main task included two single conditions with the grouping cues acting alone as well as two competing conditions displaying the grouping factors pitted against one another. Conventional aggregated analyses were combined with individual analysis and both revealed a consistent pattern of processing dominance of: (1) luminance similarity over proximity, and of (2) common region over connectedness. Interestingly, the individual analyses showed that, despite the heterogeneous responses to the single conditions, the pattern of dominance between cues was robustly homogenous among the participants in the competing conditions.

Some Time Violations Go Less Noticed: Gestalt Grouping? A Brief Report

This brief report is inspired by Bolton’s (1894, Am. J. Psychol., 6, 145–238) tick-tock phenomenon, which describes an illusionary accented grouping of isochronous, non-accented click sequences. It has repeatedly been shown that in stimulus-wise grouped sequences of an XXXOOO character (where X differs from O in terms of intensity, pitch level, or filled or unfilled duration), gap deviations between groups are more prone to go unnoticed compared to deviations within a group (e.g., Fitzgibbons et al., 1974, Percept. Psychophys., 16, 522–528.). Yet, not much is known about whether comparable anisochrony insensitivities might also occur in equal-accented sequences (XXXXX). In a same/different task setting, listeners had to detect isochrony deviations that appeared in different empty-interval locations across 800 trials within a five-pulse sequence of 250 ms interonset intervals. Findings revealed a major location dependency, with least detection accuracy for gap deviations occurring in the last interval, particularly if the 250 ms gap was lengthened rather than shortened. Results are discussed in relation to potential Gestalt grouping and Nakajima et al.’s (2014, Perception, 33, 1061–1079) perceptual assimilation and contrast observations in three-tone sequences.

Bayesian Edge Detector Using Deformable Directivity-Aware Sampling Window

Conventional image entropy merely involves the overall pixel intensity statistics which cannot respond to intensity patterns over spatial domain. However, spatial distribution of pixel intensity is definitely crucial to any biological or computer vision system, and that is why gestalt grouping rules involve using features of both aspects. Recently, the increasing integration of knowledge from gestalt research into visualization-related techniques has fundamentally altered both fields, offering not only new research questions, but also new ways of solving existing issues. This paper presents a Bayesian edge detector called GestEdge, which is effective in detecting gestalt edges, especially useful for forming object boundaries as perceived by human eyes. GestEdge is characterized by employing a directivity-aware sampling window or mask that iteratively deforms to probe or explore the existence of principal direction of sampling pixels; when convergence is reached, the window covers pixels best representing the directivity in compliance with the similarity and proximity laws in gestalt theory. During the iterative process based on the unsupervised Expectation-Minimization (EM) algorithm, the shape of the sampling window is optimally adjusted. Such a deformable window allows us to exploit the similarity and proximity among the sampled pixels. Comparisons between GestEdge and other edge detectors are shown to justify the effectiveness of GestEdge in extracting the gestalt edges.

Context-Aware Superpixel and Bilateral Entropy—Image Coherence Induces Less Entropy

Superpixel clustering is one of the most popular computer vision techniques that aggregates coherent pixels into perceptually meaningful groups, taking inspiration from Gestalt grouping rules. However, due to brain complexity, the underlying mechanisms of such perceptual rules are unclear. Thus, conventional superpixel methods do not completely follow them and merely generate a flat image partition rather than hierarchical ones like a human does. In addition, those methods need to initialize the total number of superpixels, which may not suit diverse images. In this paper, we first propose context-aware superpixel (CASP) that follows both Gestalt grouping rules and the top-down hierarchical principle. Thus, CASP enables to adapt the total number of superpixels to specific images automatically. Next, we propose bilateral entropy, with two aspects conditional intensity entropy and spatial occupation entropy, to evaluate the encoding efficiency of image coherence. Extensive experiments demonstrate CASP achieves better superpixel segmentation performance and less entropy than baseline methods. More than that, using Pearson’s correlation coefficient, a collection of data with a total of 120 samples demonstrates a strong correlation between local image coherence and superpixel segmentation performance. Our results inversely support the reliability of above-mentioned perceptual rules, and eventually, we suggest designing novel entropy criteria to test the encoding efficiency of more complex patterns.

SSNet: Learning Mid-Level Image Representation Using Salient Superpixel Network

In the standard bag-of-visual-words (BoVW) model, the burstiness problem of features and the ignorance of high-order information often weakens the discriminative power of image representation. To tackle them, we present a novel framework, named the Salient Superpixel Network, to learn the mid-level image representation. For reducing the impact of burstiness occurred in the background region, we use the salient regions instead of the whole image to extract local features, and a fast saliency detection algorithm based on the Gestalt grouping principle is proposed to generate image saliency maps. In order to introduce the high-order information, we propose a weighted second-order pooling (WSOP) method, which is capable of exploiting the high-order information and further alleviating the impact of burstiness in the foreground region. Then, we conduct experiments on six image classification benchmark datasets, and the results demonstrate the effectiveness of the proposed framework with either the handcrafted or the off-the-shelf CNN features.

Typical Utilization of Gestalt Grouping Cues in Shape Perception by Persons with Autism Spectrum Disorder

The common finding of better locally oriented perception among persons with autism spectrum disorder (ASD) is based on evidence from paradigms in which hierarchical stimuli are used to pit local and global processes against one another. However, in most cases, determining whether group differences reflect reduced global processing, enhanced local processing, or both is difficult. To provide more conclusive evidence for global perception in ASD, we examined shape formation and sensitivity to Gestalt heuristics. Children with persons with ASD and mental age matched typically developing children completed tasks in which the organization of contour segments into a shape was likely to depend on utilizing cues of closure, spatial proximity, and collinearity. In Experiment 1, search efficiency was measured, with the efficiency of the global organization indicated by the slope of the best-fitting linear reaction-time function over the number of presented items. In Experiment 2, contour integration task was administered, while Gestalt cues and the contour to background spacing ratio were manipulated independently. The findings indicated typical shape formation among the persons with ASD. Furthermore, certain interactive relations between Gestalt grouping cues that are known to govern shape formation in typically developing individuals determined the extraction of the global shape among the participants with ASD.