The psychophysics of human three-dimensional active visuospatial problem-solving.

Our understanding of how visual systems detect, analyze and interpret visual stimuli has advanced greatly. However, the visual systems of all animals do much more; they enable visual behaviours. How well the visual system performs while interacting with the visual environment and how vision is used in the real world is far from fully understood, especially in humans. It has been suggested that comparison is the most primitive of psychophysical tasks. Thus, as a probe into these active visual behaviours, we use a same-different task: Are two physical 3D objects visually the same? This task is a fundamental cognitive ability. We pose this question to human subjects who are free to move about and examine two real objects in a physical 3D space. The experimental design is such that all behaviours are directed to viewpoint change. Without any training, our participants achieved a mean accuracy of 93.82%. No learning effect was observed on accuracy after many trials, but some effect was seen for response time, number of fixations and extent of head movement. Our probe task, even though easily executed at high-performance levels, uncovered a surprising variety of complex strategies for viewpoint control, suggesting that solutions were developed dynamically and deployed in a seemingly directed hypothesize-and-test manner tailored to the specific task. Subjects need not acquire task-specific knowledge; instead, they formulate effective solutions right from the outset, and as they engage in a series of attempts, those solutions progressively refine, becoming more efficient without compromising accuracy.

Learning a Model of Shape Selectivity in V4 Cells Reveals Shape Encoding Mechanisms in the Brain.

The mechanisms involved in transforming early visual signals to curvature representations in V4 are unknown. We propose a hierarchical model that reveals V1/V2 encodings that are essential components for this transformation to the reported curvature representations in V4. Then, by relaxing the often-imposed prior of a single Gaussian, V4 shape selectivity is learned in the last layer of the hierarchy from Macaque V4 responses. We found that V4 cells integrate multiple shape parts from the full spatial extent of their receptive fields with similar excitatory and inhibitory contributions. Our results uncover new details in existing data about shape selectivity in V4 neurons that with additional experiments can enhance our understanding of processing in this area. Accordingly, we propose designs for a stimulus set that allow removing shape parts without disturbing the curvature signal to isolate part contributions to V4 responses.SIGNIFICANCE STATEMENT Selectivity to convex and concave shape parts in V4 neurons has been repeatedly reported. Nonetheless, the mechanisms that yield such selectivities in the ventral stream remain unknown. We propose a hierarchical computational model that incorporates findings of the various visual areas involved in shape processing and suggest mechanisms that transform the shape signal from low-level features to convex/concave part representations. Learning shape selectivity from Macaque V4 responses in the final processing stage in our model, we found that V4 neurons integrate shape parts from the full spatial extent of their receptive field with both facilitatory and inhibitory contributions. These results reveal hidden information in existing V4 data that with additional experiments can enhance our understanding of processing in V4.

A cortical zoom-in operation underlies covert shifts of visual spatial attention.

Shifting the focus of attention without moving the eyes poses challenges for signal coding in visual cortex in terms of spatial resolution, signal routing, and cross-talk. Little is known how these problems are solved during focus shifts. Here, we analyze the spatiotemporal dynamic of neuromagnetic activity in human visual cortex as a function of the size and number of focus shifts in visual search. We find that large shifts elicit activity modulations progressing from highest (IT) through mid-level (V4) to lowest hierarchical levels (V1). Smaller shifts cause those modulations to start at lower levels in the hierarchy. Successive shifts involve repeated backward progressions through the hierarchy. We conclude that covert focus shifts arise from a cortical coarse-to-fine process progressing from retinotopic areas with larger toward areas with smaller receptive fields. This process localizes the target and increases the spatial resolution of selection, which resolves the above issues of cortical coding.

Attention to visual motion suppresses neuronal and behavioral sensitivity in nearby feature space.

BACKGROUND: Feature-based attention prioritizes the processing of the attended feature while strongly suppressing the processing of nearby ones. This creates a non-linearity or "attentional suppressive surround" predicted by the Selective Tuning model of visual attention. However, previously reported effects of feature-based attention on neuronal responses are linear, e.g., feature-similarity gain. Here, we investigated this apparent contradiction by neurophysiological and psychophysical approaches. RESULTS: Responses of motion direction-selective neurons in area MT/MST of monkeys were recorded during a motion task. When attention was allocated to a stimulus moving in the neurons' preferred direction, response tuning curves showed its minimum for directions 60-90° away from the preferred direction, an attentional suppressive surround. This effect was modeled via the interaction of two Gaussian fields representing excitatory narrowly tuned and inhibitory widely tuned inputs into a neuron, with feature-based attention predominantly increasing the gain of inhibitory inputs. We further showed using a motion repulsion paradigm in humans that feature-based attention produces a similar non-linearity on motion discrimination performance. CONCLUSIONS: Our results link the gain modulation of neuronal inputs and tuning curves examined through the feature-similarity gain lens to the attentional impact on neural population responses predicted by the Selective Tuning model, providing a unified framework for the documented effects of feature-based attention on neuronal responses and behavior.

When We Study the Ability to Attend, What Exactly Are We Trying to Understand?

When we study the human ability to attend, what exactly do we seek to understand? It is not clear what the answer might be to this question. There is still so much to know, while acknowledging the tremendous progress of past decades of research. It is as if each new study adds a tile to the mosaic that, when viewed from a distance, we hope will reveal the big picture of attention. However, there is no map as to how each tile might be placed nor any guide as to what the overall picture might be. It is like digging up bits of mosaic tile at an ancient archeological site with no key as to where to look and then not only having to decide which picture it belongs to but also where exactly in that puzzle it should be placed. I argue that, although the unearthing of puzzle pieces is very important, so is their placement, but this seems much less emphasized. We have mostly unearthed a treasure trove of puzzle pieces but they are all waiting for cleaning and reassembly. It is an activity that is scientifically far riskier, but with great risk comes a greater reward. Here, I will look into two areas of broad agreement, specifically regarding visual attention, and dig deeper into their more nuanced meanings, in the hope of sketching a starting point for the guide to the attention mosaic. The goal is to situate visual attention as a purely computational problem and not as a data explanation task; it may become easier to place the puzzle pieces once you understand why they exist in the first place.

Early recurrence enables figure border ownership.

Rubin's face-vase illusion demonstrates how one can switch back and forth between two different interpretations depending on how the figure outlines are assigned. In the primate visual system, assigning ownership along figure borders is encoded by neurons called the border ownership (BO) cells. Studies show that the responses of these neurons not only depend on the local features within their receptive fields, but also on contextual information. Despite two decades of studies on BO neurons, the ownership assignment mechanism in the brain is still unknown. Here, we propose a hierarchical recurrent model grounded on the hypothesis that neurons in the dorsal stream provide the context required for ownership assignment. Our proposed model incorporates early recurrence from the dorsal pathway as well as lateral modulations within the ventral stream. While dorsal modulations initiate the response difference to figure on either side of the border, lateral modulations enhance the difference. We found responses of our dorsally-modulated BO cells, similar to their biological counterparts, are invariant to size, position and solid/outlined figures. Moreover, our model BO cells exhibit comparable levels of reliability in the ownership signal to biological BO neurons. We found dorsal modulations result in high levels of accuracy and robustness for BO assignments in complex scenes compared to previous models based on ventral feedback. Finally, our experiments with illusory contours suggest that BO encoding could explain the perception of such contours in higher processing stages in the brain.

On the control of attentional processes in vision.

The study of attentional processing in vision has a long and deep history. Recently, several papers have presented insightful perspectives into how the coordination of multiple attentional functions in the brain might occur. These begin with experimental observations and the authors propose structures, processes, and computations that might explain those observations. Here, we consider a perspective that past works have not, as a complementary approach to the experimentally-grounded ones. We approach the same problem as past authors but from the other end of the computational spectrum, from the problem nature, as Marr's Computational Level would prescribe. What problem must the brain solve when orchestrating attentional processes in order to successfully complete one of the myriad possible visuospatial tasks at which we as humans excel? The hope, of course, is for the approaches to eventually meet and thus form a complete theory, but this is likely not soon. We make the first steps towards this by addressing the necessity of attentional control, examining the breadth and computational difficulty of the visuospatial and attentional tasks seen in human behavior, and suggesting a sketch of how attentional control might arise in the brain. The key conclusions of this paper are that an executive controller is necessary for human attentional function in vision, and that there is a 'first principles' computational approach to its understanding that is complementary to the previous approaches that focus on modelling or learning from experimental observations directly.

Long-term memory and hippocampal function support predictive gaze control during goal-directed search.

Eye movements during visual search change with prior experience for search stimuli. Previous studies measured these gaze effects shortly after initial viewing, typically during free viewing; it remains open whether the effects are preserved across long delays and for goal-directed search, and which memory system guides gaze. In Experiment 1, we analyzed eye movements of healthy adults viewing novel and repeated scenes while searching for a scene-embedded target. The task was performed across different time points to examine the repetition effects in long-term memory, and memory types were grouped based on explicit recall of targets. In Experiment 2, an amnesic person with bilateral extended hippocampal damage and the age-matched control group performed the same task with shorter intervals to determine whether or not the repetition effects depend on hippocampal function. When healthy adults explicitly remembered repeated target-scene pairs, search time and fixation duration decreased, and gaze was directed closer to the target region, than when they forgot targets. These effects were seen even after a one-month delay from their initial viewing, suggesting the effects are associated with long-term, explicit memory. Saccadic amplitude was not strongly modulated by scene repetition or explicit recall of targets. The amnesic person did not show explicit recall or implicit repetition effects, whereas his control group showed similar patterns to those seen in Experiment 1. The results reveal several aspects of gaze control that are influenced by long-term memory. The dependence of gaze effects on medial temporal lobe integrity support a role for this region in predictive gaze control.

Multiplicative modulations enhance diversity of hue-selective cells.

There is still much to understand about the brain's colour processing mechanisms and the transformation from cone-opponent representations to perceptual hues. Moreover, it is unclear which area(s) in the brain represent unique hues. We propose a hierarchical model inspired by the neuronal mechanisms in the brain for local hue representation, which reveals the contributions of each visual cortical area in hue representation. Hue encoding is achieved through incrementally increasing processing nonlinearities beginning with cone input. Besides employing nonlinear rectifications, we propose multiplicative modulations as a form of nonlinearity. Our simulation results indicate that multiplicative modulations have significant contributions in encoding of hues along intermediate directions in the MacLeod-Boynton diagram and that our model V2 neurons have the capacity to encode unique hues. Additionally, responses of our model neurons resemble those of biological colour cells, suggesting that our model provides a novel formulation of the brain's colour processing pathway.

Incremental Learning Through Deep Adaptation.

Given an existing trained neural network, it is often desirable to learn new capabilities without hindering performance of those already learned. Existing approaches either learn sub-optimal solutions, require joint training, or incur a substantial increment in the number of parameters for each added domain, typically as many as the original network. We propose a method called Deep Adaptation Modules (DAM) that constrains newly learned filters to be linear combinations of existing ones. DAMs precisely preserve performance on the original domain, require a fraction (typically 13 percent, dependent on network architecture) of the number of parameters compared to standard fine-tuning procedures and converge in less cycles of training to a comparable or better level of performance. When coupled with standard network quantization techniques, we further reduce the parameter cost to around 3 percent of the original with negligible or no loss in accuracy. The learned architecture can be controlled to switch between various learned representations, enabling a single network to solve a task from multiple different domains. We conduct extensive experiments showing the effectiveness of our method on a range of image classification tasks and explore different aspects of its behavior.

Correction: Rapid visual categorization is not guided by early salience-based selection.

[This corrects the article DOI: 10.1371/journal.pone.0224306.].

Rapid visual categorization is not guided by early salience-based selection.

The current dominant visual processing paradigm in both human and machine research is the feedforward, layered hierarchy of neural-like processing elements. Within this paradigm, visual saliency is seen by many to have a specific role, namely that of early selection. Early selection is thought to enable very fast visual performance by limiting processing to only the most salient candidate portions of an image. This strategy has led to a plethora of saliency algorithms that have indeed improved processing time efficiency in machine algorithms, which in turn have strengthened the suggestion that human vision also employs a similar early selection strategy. However, at least one set of critical tests of this idea has never been performed with respect to the role of early selection in human vision. How would the best of the current saliency models perform on the stimuli used by experimentalists who first provided evidence for this visual processing paradigm? Would the algorithms really provide correct candidate sub-images to enable fast categorization on those same images? Do humans really need this early selection for their impressive performance? Here, we report on a new series of tests of these questions whose results suggest that it is quite unlikely that such an early selection process has any role in human rapid visual categorization.

Feed-forward visual processing suffices for coarse localization but fine-grained localization in an attention-demanding context needs feedback processing.

It is well known that simple visual tasks, such as object detection or categorization, can be performed within a short period of time, suggesting the sufficiency of feed-forward visual processing. However, more complex visual tasks, such as fine-grained localization may require high-resolution information available at the early processing levels in the visual hierarchy. To access this information using a top-down approach, feedback processing would need to traverse several stages in the visual hierarchy and each step in this traversal takes processing time. In the present study, we compared the processing time required to complete object categorization and localization by varying presentation duration and complexity of natural scene stimuli. We hypothesized that performance would be asymptotic at shorter presentation durations when feed-forward processing suffices for visual tasks, whereas performance would gradually improve as images are presented longer if the tasks rely on feedback processing. In Experiment 1, where simple images were presented, both object categorization and localization performance sharply improved until 100 ms of presentation then it leveled off. These results are a replication of previously reported rapid categorization effects but they do not support the role of feedback processing in localization tasks, indicating that feed-forward processing enables coarse localization in relatively simple visual scenes. In Experiment 2, the same tasks were performed but more attention-demanding and ecologically valid images were used as stimuli. Unlike in Experiment 1, both object categorization performance and localization precision gradually improved as stimulus presentation duration became longer. This finding suggests that complex visual tasks that require visual scrutiny call for top-down feedback processing.

Development of spatial suppression surrounding the focus of visual attention.

In adulthood, research has demonstrated that surrounding the spatial location of attentional focus is a suppressive field, resulting from top-down attention promoting the processing of relevant stimuli and inhibiting surrounding distractors (e.g., Hopf et al., 2006). It is not fully known, however, how this phenomenon manifests during development. This is an important question since attention processes are likely even more critical in development because of their potential impact on learning and day-to-day activities. The current study examined whether spatial suppression surrounding the focus of visual attention, a predicted by-product of top-down attentional modulation, is observed in development. A wide age range separated in six incremental age levels was included, allowing for a detailed examination of potential differences in the effect of attention on visual processing across development. Participants between 12 and 27 years of age exhibited spatial suppression surrounding their focus of visual attention. Their accuracy increased as a function of the separation distance between a spatially cued (and attended) target and a second target, suggesting that a ring of suppression surrounded the attended target. Attentional surround suppression was not observed in 8- to 11-years-olds, even with a longer spatial cue presentation time, demonstrating that the lack of the effect at these ages is not due to slowed attentional feedback processes. Our findings demonstrate that top-down attentional processes exhibit functional maturity beginning around 12 years of age with continuing maturation of their expression until 17, which likely impacts education and the diagnosis of visual and cognitive clinical pathologies.

Attention: The Messy Reality.

The human capability to attend has been both considered as easy and as impossible to understand by philosophers and scientists through the centuries. Much has been written by brain, cognitive, and philosophical scientists trying to explain attention as it applies to sensory and reasoning processes, let alone consciousness. It has been only in the last few decades that computational scientists have entered the picture adding a new language with which to express attentional behavior and function. This new perspective has produced some progress to the centuries-old goal, but there is still far to go. Although a central belief in many scientific disciplines has been to seek a unifying explanatory principle for natural observations, it may be that we need to put this aside as it applies to attention and accept the fact that attention is really an integrated set of mechanisms, too messy to cleanly and parsimoniously express with a single principle. These mechanisms are claimed to be critical to enable functional generalization of brain processes and thus an integrative perspective is important. Here we present first steps towards a theoretical and algorithmic view on how the many different attentional mechanisms may be deployed, coordinated, synchronized, and effectively utilized. A hierarchy of dynamically defined closed-loop control processes is proposed, each with its own optimization objective, which is extensible to multiple layers. Although mostly speculative, simulation and experimental work support important components.

The Attentional Suppressive Surround: Eccentricity, Location-Based and Feature-Based Effects and Interactions.

The Selective Tuning model of visual attention (Tsotsos, 1990) has proposed that the focus of attention is surrounded by an inhibitory zone, eliciting a center-surround attentional distribution. This attentional suppressive surround inhibits irrelevant information which is located close to attended information in physical space (e.g., Cutzu and Tsotsos, 2003; Hopf et al., 2010) or in feature space (e.g., Tombu and Tsotsos, 2008; Störmer and Alvarez, 2014; Bartsch et al., 2017). In Experiment 1, we investigate the interaction between location-based and feature-based surround suppression and hypothesize that the attentional surround suppression would be maximized when spatially adjacent stimuli are also represented closely within a feature map. Our results demonstrate that perceptual discrimination is worst when two similar orientations are presented in proximity to each other, suggesting the interplay of the two surround suppression mechanisms. The Selective Tuning model also predicts that the size of the attentional suppressive surround is determined by the receptive field size of the neuron which optimally processes the attended information. The receptive field size of the processing neurons is tightly associated with stimulus size and eccentricity. Therefore, Experiment 2 tested the hypothesis that the size of the attentional suppressive surround would become larger as stimulus size and eccentricity increase, corresponding to an increase in the neuron's receptive field size. We show that stimulus eccentricity but not stimulus size modulates the size of the attentional suppressive surround. These results are consistent for both low- and high-level features (e.g., orientation and human faces). Overall, the present study supports the existence of the attentional suppressive surround and reveals new properties of this selection mechanism.

Visual attention and its intimate links to spatial cognition.

It is almost universal to regard attention as the facility that permits an agent, human or machine, to give priority processing resources to relevant stimuli while ignoring the irrelevant. The reality of how this might manifest itself throughout all the forms of perceptual and cognitive processes possessed by humans, however, is not as clear. Here, we examine this reality with a broad perspective in order to highlight the myriad ways that attentional processes impact both perception and cognition. The paper concludes by showing two real-world problems that exhibit sufficient complexity to illustrate the ways in which attention and cognition connect. These then point to new avenues of research that might illuminate the overall cognitive architecture of spatial cognition.

Short and Long-Term Attentional Firing Rates Can Be Explained by ST-Neuron Dynamics.

Attention modulates neural selectivity and optimizes the allocation of cortical resources during visual tasks. A large number of experimental studies in primates and humans provide ample evidence. As an underlying principle of visual attention, some theoretical models suggested the existence of a gain element that enhances contrast of the attended stimuli. In contrast, the Selective Tuning model of attention (ST) proposes an attentional mechanism based on suppression of irrelevant signals. In this paper, we present an updated characterization of the ST-neuron proposed by the Selective Tuning model, and suggest that the inclusion of adaptation currents (Ih) to ST-neurons may explain the temporal profiles of the firing rates recorded in single V4 cells during attentional tasks. Furthermore, using the model we show that the interaction between stimulus-selectivity of a neuron and attention shapes the profile of the firing rate, and is enough to explain its fast modulation and other discontinuities observed, when the neuron responds to a sudden switch of stimulus, or when one stimulus is added to another during a visual task.

Revisiting active perception.

Despite the recent successes in robotics, artificial intelligence and computer vision, a complete artificial agent necessarily must include active perception. A multitude of ideas and methods for how to accomplish this have already appeared in the past, their broader utility perhaps impeded by insufficient computational power or costly hardware. The history of these ideas, perhaps selective due to our perspectives, is presented with the goal of organizing the past literature and highlighting the seminal contributions. We argue that those contributions are as relevant today as they were decades ago and, with the state of modern computational tools, are poised to find new life in the robotic perception systems of the next decade.

Attention to Color Sharpens Neural Population Tuning via Feedback Processing in the Human Visual Cortex Hierarchy.

Attention can facilitate the selection of elementary object features such as color, orientation, or motion. This is referred to as feature-based attention and it is commonly attributed to a modulation of the gain and tuning of feature-selective units in visual cortex. Although gain mechanisms are well characterized, little is known about the cortical processes underlying the sharpening of feature selectivity. Here, we show with high-resolution magnetoencephalography in human observers (men and women) that sharpened selectivity for a particular color arises from feedback processing in the human visual cortex hierarchy. To assess color selectivity, we analyze the response to a color probe that varies in color distance from an attended color target. We find that attention causes an initial gain enhancement in anterior ventral extrastriate cortex that is coarsely selective for the target color and transitions within ∼100 ms into a sharper tuned profile in more posterior ventral occipital cortex. We conclude that attention sharpens selectivity over time by attenuating the response at lower levels of the cortical hierarchy to color values neighboring the target in color space. These observations support computational models proposing that attention tunes feature selectivity in visual cortex through backward-propagating attenuation of units less tuned to the target.SIGNIFICANCE STATEMENT Whether searching for your car, a particular item of clothing, or just obeying traffic lights, in everyday life, we must select items based on color. But how does attention allow us to select a specific color? Here, we use high spatiotemporal resolution neuromagnetic recordings to examine how color selectivity emerges in the human brain. We find that color selectivity evolves as a coarse to fine process from higher to lower levels within the visual cortex hierarchy. Our observations support computational models proposing that feature selectivity increases over time by attenuating the responses of less-selective cells in lower-level brain areas. These data emphasize that color perception involves multiple areas across a hierarchy of regions, interacting with each other in a complex, recursive manner.