Article

Neural Mechanisms of Selective Visual Attention

Authors:
To read the full-text of this research, you can request a copy directly from the author.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... In the cortex, long-range projections from the frontal area (i.e., top) innervate and regulate neuronal activity and information processing in the sensory area (i.e., bottom). Such top-down input can actively modulate sensory experience and support diverse perceptual and cognitive behaviors, such as attention, value, and memory (Buschman and Miller, 2014;Corbetta and Shulman, 2002;Desimone and Duncan, 1995;Gilbert and Sigman, 2007;Miller and Cohen, 2001;Ruff, 2013;Tomita et al., 1999;Tsodyks and Gilbert, 2004). For example, in the mouse somatosensory cortex, top-down projection from the motor cortex promotes large calcium signals and persistent firing of layer 5 pyramidal neurons and facilitates accurate perception of tactile surfaces (Manita et al., 2015;Sreenivasan et al., 2016;Xu et al., 2012). ...
... Similarly, projection from the motor cortex to the auditory cortex (AUD) has been found to regulate the activities of auditory cortical neurons Schneider et al., 2014). Top-down regulation is a powerful mechanism for focusing on behaviorally relevant information and filtering out irrelevant information (Buschman and Miller, 2014;Corbetta and Shulman, 2002;Desimone and Duncan, 1995;Miller and Cohen, 2001;Ruff, 2013). Although recent virus-assisted circuit mapping and optogenetic manipulation have provided important insights into top-down interaction between the frontal and sensory areas of the cortex at the population level (DeNardo et al., 2015;Miyamoto et al., 2016;Nelson et al., 2013;Schneider et al., 2014;Sreenivasan et al., 2016;Zhang et al., 2014Zhang et al., , 2016Zhang et al., 2014), the precise organization of the underlying neural circuits remains poorly understood. ...
... Long-range input from the frontal area to the sensory area actively modulates how sensory information is analyzed in the cortex. Therefore, it is crucial for information processing and complex behavior (Corbetta and Shulman, 2002;Desimone and Duncan, 1995;Miller and Cohen, 2001;Ruff, 2013). Previous studies have revealed the reciprocal innervations, largely at the (legend continued on next page) population level, between the frontal and sensory areas of the cortex (Aronoff et al., 2010;DeNardo et al., 2015;Manita et al., 2015;Mao et al., 2011;Miyamoto et al., 2016;Xu et al., 2012;Zhang et al., 2014Zhang et al., , 2016Zingg et al., 2014). ...
Article
The frontal area of the cerebral cortex provides long-range inputs to sensory areas to modulate neuronal activity and information processing. These long-range circuits are crucial for accurate sensory perception and complex behavioral control; however, little is known about their precise circuit organization. Here we specifically identified the presynaptic input neurons to individual excitatory neuron clones as a unit that constitutes functional microcircuits in the mouse sensory cortex. Interestingly, the long-range input neurons in the frontal but not contralateral sensory area are spatially organized into discrete vertical clusters and preferentially form synapses with each other over nearby non-input neurons. Moreover, the assembly of distant presynaptic microcircuits in the frontal area depends on the selective synaptic communication of excitatory neuron clones in the sensory area that provide inputs to the frontal area. These findings suggest that highly precise long-range reciprocal microcircuit-to-microcircuit communication mediates frontal-sensory area interactions in the mammalian cortex.
... A second explanation for the discrepancy is that the abundance of recurrent connections in cortex belies a superficial role in neural computation. Perhaps the core computations can be performed by a feedforward network [34], while recurrent processing serves more auxiliary and modulatory functions, such as divisive normalization [35] and attention [36][37][38][39]. This perspective is convenient because it enables us to hold on to the feedforward model in our minds. ...
... Such top-down effects have been grouped under the label 'attention', and they have been the subject of an entire subfield of study. For our purposes, it is sufficient to note that the effects and mechanisms of top-down attention are well-documented and pervasive in visual cortex (for review, see [36][37][38]), and thus there is no question that this is one important function of recurrent connections. ...
... The winning hypothesis could be the maximum a posteriori (MAP) hypothesis or the maximum likelihood hypothesis. Influential models of visual inference involving competitive recurrent interactions include divisive normalization [35], biased competition [36], and predictive coding [30,32,77]. ...
Article
Full-text available
Biological visual systems exhibit abundant recurrent connectivity. State-of-the-art neural network models for visual recognition, by contrast, rely heavily or exclusively on feedforward computation. Any finite-time recurrent neural network (RNN) can be unrolled along time to yield an equivalent feedforward neural network (FNN). This important insight suggests that computational neuroscientists may not need to engage recurrent computation, and that computer-vision engineers may be limiting themselves to a special case of FNN if they build recurrent models. Here we argue, to the contrary, that FNNs are a special case of RNNs and that computational neuroscientists and engineers should engage recurrence to understand how brains and machines can (1) achieve greater and more flexible computational depth (2) compress complex computations into limited hardware (3) integrate priors and priorities into visual inference through expectation and attention (4) exploit sequential dependencies in their data for better inference and prediction and (5) leverage the power of iterative computation.
... These functions modulate incoming sensory signals and influence their processing at multiple stages to inform awareness, decisions, actions, and subsequent memories. It has long been appreciated that short-term memory, or ''working memory'' (WM), plays an important role in forming attentional templates (e.g., Desimone and Duncan, 1995). Here we suggest that WM is part of a much larger family of heterogenous attention-guiding memory traces that span multiple timescales. ...
... At slightly longer timeframes, WM provides a limited set of more durable traces that are independent of continuous sensory stimulation and resistant to interference and that act to guide adaptive behavior (Baddeley, 2003). The fundamental role WM plays in guiding attention is widely recognized and has been studied extensively (Desimone and Duncan, 1995). Information in WM has been considered the major source of top-down proactive attention. ...
... The classic biased-competition model of attention proposed that perceptual templates in WM bias visual processing to prioritize task-relevant items (Desimone and Duncan, 1995). This idea is grounded in the notion that WM is maintained via persistent activation of sensory-specific neural codes (e.g., Chelazzi et al., 1993), resulting in an elevated baseline for subsequent processing of related input. ...
Article
Full-text available
Memories are about the past, but they serve the future. Memory research often emphasizes the former aspect: focusing on the functions that re-constitute (re-member) experience and elucidating the various types of memories and their interrelations, timescales, and neural bases. Here we highlight the prospective nature of memory in guiding selective attention, focusing on functions that use previous experience to anticipate the relevant events about to unfold-to "premember" experience. Memories of various types and timescales play a fundamental role in guiding perception and performance adaptively, proactively, and dynamically. Consonant with this perspective, memories are often recorded according to expected future demands. Using working memory as an example, we consider how mnemonic content is selected and represented for future use. This perspective moves away from the traditional representational account of memory toward a functional account in which forward-looking memory traces are informationally and computationally tuned for interacting with incoming sensory signals to guide adaptive behavior.
... Anyone who has tried to find a particular product in a supermarket aisle has encountered that visual search is limited by what we can attend to at any moment in time. These limitations are part and parcel of theories of visual search and of selective attention in general (Bundesen, 1990;Bundesen et al., 2005;Desimone & Duncan, 1995;Huang & Pashler, 2007;Treisman, 1988;Treisman & Gelade, 1980;Wolfe, 1994Wolfe, , 2007. However, the same limitations are also central to an important and currently much debated question: Can individuals look for multiple objects at the same time? ...
... Kristjánsson & Campana, 2010;Maljkovic & Nakayama, 1994;Urai et al., 2019). However, as the goal of visual search typically changes from situation to situation, in the current review we will focus on a third source of bias, namely top-down, goal-directed mechanisms, which establish the flexible, endogenous biases that enable the observer to adapt to dynamically changing task goals (Baluch & Itti, 2011;Chelazzi et al., 1993;Desimone & Duncan, 1995). Figure 1(a) presents what we regard as the canonical, core model of such goal-driven visual search (Chelazzi et al., 1993;Desimone & Duncan, 1995;Duncan & Humphreys, 1989;Eimer, 2014;Huang & Pashler, 2007;Wolfe, 1994). 1 First, prior to search, during the preparation stage, the observer activates a memory representation of the target object or its most important defining feature(s)for example red and round when looking for tomatoes. ...
... However, as the goal of visual search typically changes from situation to situation, in the current review we will focus on a third source of bias, namely top-down, goal-directed mechanisms, which establish the flexible, endogenous biases that enable the observer to adapt to dynamically changing task goals (Baluch & Itti, 2011;Chelazzi et al., 1993;Desimone & Duncan, 1995). Figure 1(a) presents what we regard as the canonical, core model of such goal-driven visual search (Chelazzi et al., 1993;Desimone & Duncan, 1995;Duncan & Humphreys, 1989;Eimer, 2014;Huang & Pashler, 2007;Wolfe, 1994). 1 First, prior to search, during the preparation stage, the observer activates a memory representation of the target object or its most important defining feature(s)for example red and round when looking for tomatoes. This memory representation is typically referred to as the search template, attentional set, or attentional control settings (henceforth simply template). ...
Article
Full-text available
Can individuals look for multiple objects at the same time? A simple question, but answering it has proven difficult. In this review, we describe possible cognitive architectures and their predictions about the capacity of visual search. We broadly distinguish three stages at which limitations may occur: (1) preparation (establishing and maintaining a mental representation of a search target), (2) selection (using this mental representation to extract candidate targets from the visual input), and (3) post-selection processing (verifying that the selected information actually is a target). We then review the empirical evidence from various paradigms, together with their strengths and pitfalls. The emerging picture is that multiple target search comes with costs, but the magnitude of this cost differs depending on the processing stage. Selection appears strongly limited, while preparation of multiple search target representations in anticipation of a search is possible with relatively small costs. Finally, there is currently not sufficient information to determine the capacity limitations of post-selection processing. We hope that our review contributes to better targeted research into the mechanisms of multiple-target search. A better understanding of multiple-target search will also contribute to better design of real-life multiple-target search problems, reducing the risk of detrimental search failures.
... However, more recent models of intelligence have included attention control as a primary factor contributing to individual differences in fluid intelligence (Conway et al., 1999(Conway et al., , 2003Heitz et al., 2005;Kovacs & Conway, 2016;Kyllonen & Christal, 1990;Shipstead et al., 2016). And more recent neural models of perception have proposed attention as an important mechanism at early and late stages of perception (Deco & Rolls, 2005;Desimone & Duncan, 1995;Kok, 1997;Luck et al., 2000;O'Craven et al., 1997;Usher & Niebur, 1996). Therefore, attention processes may play an important role in relating lower-level sensory abilities to higher-order cognitive abilities. ...
... In fact, much of the early cognitive neuroscience work focused on understanding the effects of top-down attention on early perceptual processing. This research lead to the understanding that attention leads to the enhancement and sensitivity of neuronal response in early perceptual brain regions to bias neural competition towards goal-relevant behavior (Desimone & Duncan, 1995;Reynolds & Pasternak, 2000). Additionally, recent work by Unsworth and colleagues has suggested that the voluntary control of the intensity of attention is important for understanding the relationship between working-memory capacity and attention control (A. ...
... The role of attention on early and late perceptual stages of processing is well studied in the neurophysiological and neuroimaging literatures (Deco & Rolls, 2005;Kok, 1997;Luck et al., 2000;O'Craven et al., 1997;Usher & Niebur, 1996). One of the most influential theories, biased competition theory, states that attention to one stimulus or stimulus feature biases the neural activation in neurons that respond to that stimulus or stimulus feature (Desimone & Duncan, 1995). This bias is not only observed by an increased activation in the relevant brain areas but an inhibition of activity in nearby non-relevant brain areas, thereby increasing the signal-to-noise ratio of attended stimuli over unattended stimuli. ...
Article
Full-text available
Intelligence is correlated with the ability to make fine sensory discriminations. Although this relationship has been known since the beginning of intelligence testing, the mechanisms underlying this relationship are still unknown. In two large-scale structural equation-modelling studies, we investigated whether individual differences in attention control abilities can explain the relationship between sensory discrimination and intelligence. Across these two studies, we replicated the finding that attention control fully mediated the relationships of intelligence/working-memory capacity to sensory discrimination. Our findings show that attention control plays a prominent role in relating sensory discrimination to higher-order cognitive abilities.
... Like most vertebrates, humans can only obtain a part of their visual field at a high acuity and therefore repeatedly move their eyes in order to construct a representation of their environment with sufficiently high resolution (Land & Fernald, 1992). Controlling gaze along with retrieving and filtering relevant signals from the environment is a central task of the attentional system (Desimone & Duncan, 1995). In the past, various lines of research have addressed the mechanisms driving such attentional control. ...
... Participants furthermore modulated their gaze behavior along the course of the experiment: While the first trials were characterized by a stronger tendency to gaze at the agent's face -especially when she smiled at the participants -participants more strongly prioritized gazing at the objects during the second half of the experiment. This behavioral pattern may be explained by a general bias towards novelty in visual attention (Desimone & Duncan, 1995), but may also in part result from complying to the civil inattention norm: Looking at a conspecific's face at the first encounter may be a sensible social behavior, while repeated looking at the conspecific in the absence of a conversation could be considered inadequate starring. ...
Thesis
Social attention is a ubiquitous, but also enigmatic and sometimes elusive phenomenon. We direct our gaze at other human beings to see what they are doing and to guess their intentions, but we may also absorb social events en passant as they unfold in the corner of the eye. We use our gaze as a discrete communication channel, sometimes conveying pieces of information which would be difficult to explicate, but we may also find ourselves avoiding eye-contact with others in moments when self-disclosure is fear-laden. We experience our gaze as the most genuine expression of our will, but research also suggests considerable levels of predictability and automaticity in our gaze behavior. The phenomenon’s complexity has hindered researchers from developing a unified framework which can conclusively accommodate all of its aspects, or from even agreeing on the most promising research methodologies. The present work follows a multi-methods approach, taking on several aspects of the phenomenon from various directions. Participants in study 1 viewed dynamic social scenes on a computer screen. Here, low-level physical saliency (i.e. color, contrast, or motion) and human heads both attracted gaze to a similar extent, providing a comparison of two vastly different classes of gaze predictors in direct juxtaposition. In study 2, participants with varying degrees of social anxiety walked in a public train station while their eye movements were tracked. With increasing levels of social anxiety, participants showed a relative avoidance of gaze at near compared to distant people. When replicating the experiment in a laboratory situation with a matched participant group, social anxiety did not modulate gaze behavior, fueling the debate around appropriate experimental designs in the field. Study 3 employed virtual reality (VR) to investigate social gaze in a complex and immersive, but still highly controlled situation. In this situation, participants exhibited a gaze behavior which may be more typical for real-life compared to laboratory situations as they avoided gaze contact with a virtual conspecific unless she gazed at them. This study provided important insights into gaze behavior in virtual social situations, helping to better estimate the possible benefits of this new research approach. Throughout all three experiments, participants showed consistent inter-individual differences in their gaze behavior. However, the present work could not resolve if these differences are linked to psychologically meaningful traits or if they instead have an epiphenomenal character.
... Perhaps the core computations can be performed by a feedforward network 34 , while recurrent processing serves more auxiliary and modulatory functions, such as divisive normalization 35 and attention [36][37][38][39] . This perspective is convenient because it enables us to hold on to the feedforward model in our minds. ...
... Such top-down effects have been grouped under the label "attention", and they have been the subject of an entire sub-field of study. For our purposes, it is sufficient to note that the effects and mechanisms of top-down attention are well-documented and pervasive in visual cortex (for review, see [36][37][38]), and thus there is no question that this is one important function of recurrent connections. ...
Preprint
Full-text available
Biological visual systems exhibit abundant recurrent connectivity. State-of-the-art neural network models for visual recognition, by contrast, rely heavily or exclusively on feedforward computation. Any finite-time recurrent neural network (RNN) can be unrolled along time to yield an equivalent feedforward neural network (FNN). This important insight suggests that computational neuroscientists may not need to engage recurrent computation, and that computer-vision engineers may be limiting themselves to a special case of FNN if they build recurrent models. Here we argue, to the contrary, that FNNs are a special case of RNNs and that computational neuroscientists and engineers should engage recurrence to understand how brains and machines can (1) achieve greater and more flexible computational depth, (2) compress complex computations into limited hardware, (3) integrate priors and priorities into visual inference through expectation and attention, (4) exploit sequential dependencies in their data for better inference and prediction, and (5) leverage the power of iterative computation.
... Attention is a mechanism that enables the brain to optimize performance given metabolic constraints [3]. Covert spatial attention enhances stimulus processing at relevant locations in the visual field at the expense of stimuli at other locations via a push-pull mechanism [4]-improving visual processing at the attended location (benefits) and suppressing signals at unattended locations (costs) [e.g., [5][6][7]. There are two types of covert spatial attention [1,2]. ...
... TMS should alter gain modulations at the attended and unattended locations, given that attention is selective and considered to operate via gain control [7,8,16,17,49]. In our study, given that stimulation occurred after attentional deployment, in the target-stimulated condition TMS extinguished attentional benefits and weakened attentional costs, providing converging Report evidence for the push-pull mechanism between the attended and unattended locations [4][5][6]. ...
Article
Orienting covert exogenous (involuntary) attention to a target location improves performance in many visual tasks [1 • Carrasco M. Visual attention: the past 25 years.Vision Res. 2011; 51: 1484-1525 • Crossref • PubMed • Scopus (914) • Google Scholar , 2 • Carrasco M. Spatial covert attention: Perceptual modulation. The Oxford Handbook of Attention. 183. 2014: 230 • Google Scholar ]. It is unknown whether early visual cortical areas are necessary for this improvement. To establish a causal link between these areas and attentional modulations, we used transcranial magnetic stimulation (TMS) to briefly alter cortical excitability and determine whether early visual areas mediate the effect of exogenous attention on performance. Observers performed an orientation discrimination task. After a peripheral valid, neutral, or invalid cue, two cortically magnified gratings were presented, one in the stimulated region and the other in the symmetric region in the opposite hemifield. Observers received two successive TMS pulses around their occipital pole while the stimuli were presented. Shortly after, a response cue indicated the grating whose orientation observers had to discriminate. The response cue either matched—target stimulated—or did not match—distractor stimulated—the stimulated side. Grating contrast was varied to measure contrast response functions (CRF) for all combinations of attention and TMS conditions. When the distractor was stimulated, exogenous attention yielded response gain—performance benefits in the valid-cue condition and costs in the invalid-cue condition compared with the neutral condition at the high contrast levels. Crucially, when the target was stimulated, this response gain was eliminated. Therefore, TMS extinguished the effect of exogenous attention. These results establish a causal link between early visual areas and the modulatory effect of exogenous attention on performance.
... A CNN typically contains more than seven layers number that would be almost prohibitive for a fully connected network. CNN has obtained excellent results in visual recognition tasks applied to different areas as facial expressions recognition, digit classification [12,16], satellite image classification [17], semantic image segmentation [16,18] and object classification [19]. ...
... It has been demonstrated that using the maximum value from the block (max pooling), it is more efficient in order to summarize region characteristics [18]. Max pooling operations has similarities with the biological behaviour of visual cortex and how it is capable of summarize information [19]. Max pooling operation causes a reduction in data size in a factor equal to the size of the sample window on which the operation has been developed. ...
Article
Full-text available
X-ray methods have proven to be reliable, accurate and sensitive techniques to study activated carbons. The studying of granular activated carbon (GAC) samples through X-ray digital radiographic images using Deep Learning, more specifically convolutional neural networks (CNN) class of model, has been explored. Results were compared to hand-engineered characterization using X-Ray absorption method (XRA). It was proved that CNNs represent a fast and reliable analytical tool for indirect information on the chemical and physical characteristics of GACs. The proposed method opens possibilities for the application of Deep Learning based models on radiographic images for the characterization and comparison of exhausted and virgin porous materials.
... Directing attention describes the process of prioritising and preferentially processing sensory input (Desimone and Duncan, 1995;Harris and Thiele, 2011). ...
... Attention can be shifted towards a spatial location (Posner, 1980), an object (Duncan, 1984) or feature (Rossi and Paradiso, 1995). Selective attention refers to the filtering of behavioural irrelevant information that is an essential step between perception and behavioural action (Desimone and Duncan, 1995;Driver, 2001;Johnston and Dark, 1986). Auditory attention can be voluntarily directed towards a sound object in a scene (endogenous, central, top-down). ...
Thesis
Full-text available
Natural sensory scenes are often very complex, with a multitude of overlapping objects in space and time. In order to direct behaviour, a critical aspect of everyday perception is the segregation and grouping of relevant features from those scenes, known as figure-ground segregation. The neurobiological basis of auditory figure-ground processing is poorly understood. To gain insights into different aspects of this process, I have investigated the behavioural, systemic and neuronal mechanisms the brain uses to segregate and group temporally coherent elements from a complex acoustic scene in macaque monkeys. This thesis presents the result of this research in five chapters: Chapter 1 reviews the fundamental basics of auditory scene analysis and the auditory system. Chapter 2, 3 and 4 present experimental work and cover figure detection behaviour (Chapter 2), systemic organisation of figure-ground analysis (Chapter 3) and the underlying neuronal mechanisms (Chapter 4). Finally, Chapter 5 discusses and interprets the results in the context of previous research. In summary, this work establishes that macaques are an excellent animal model for auditory scene analysis and provides new evidence of the cortical response mechanisms during auditory figure-ground segregation. I show that macaques have not only similar detection performance to humans but that the areal organisation measured with fMRI is comparable. Furthermore, I demonstrate robust effects on neuronal firing rates in response to auditory figures across the cortical hierarchy. Lastly, this thesis establishes neuronal differences in figure processing between anterior and posterior auditory cortical fields.
... In primates, the lateral prefrontal cortex (LPFC) is strongly implicated in executive control (Desimone and Duncan, 1995;Fuster, 2001;Goldman-Rakic, 2011;Miller and Cohen, 2001;Petrides, 2005), including working memory, attention to stimulus attributes, and the selection of context-dependent responses. How these functions arise from the interactions between LPFC and the rest of the brain remains unclear. ...
Article
Full-text available
The lateral prefrontal cortex (LPFC) of primates plays an important role in executive control, but how it interacts with the rest of the cortex remains unclear. To address this, we densely mapped the cortical connectome of LPFC, using electrical microstimulation combined with functional MRI (EM-fMRI). We found isomorphic mappings between LPFC and five major processing domains composing most of the cerebral cortex except early sensory and motor areas. An LPFC grid of ∼200 stimulation sites topographically mapped to separate grids of activation sites in the five domains, coarsely resembling how the visual cortex maps the retina. The temporal and parietal maps largely overlapped in LPFC, suggesting topographically organized convergence of the ventral and dorsal streams, and the other maps overlapped at least partially. Thus, the LPFC contains overlapping, millimeter-scale maps that mirror the organization of major cortical processing domains, supporting LPFC’s role in coordinating activity within and across these domains.
... Vision is the main source of sensory information in humans. Considering the abundance of visual 37 information bombarding the brain at any given moment, visual processing is inherently 38 competitive and requires neural mechanisms that suppress redundant or distracting input 39 (Desimone and Duncan, 1995). It has been suggested that, for efficient information processing, 40 the brain applies a so called 'sparse coding' strategy, i.e. encoding of sensory information using 41 small number of active neurons (Baddeley, et al., 1997;Barlow, 1961;Field, 1987; Olshausen and 42 Field, 2004). ...
Preprint
Full-text available
Spatial suppression (SS) is a visual perceptual phenomenon that is manifest in a reduction of directional sensitivity for drifting high-contrast gratings whose size exceeds the center of the visual field. Gratings moving at faster velocities induce stronger SS. The neural processes that give rise to such size- and velocity-dependent reductions in directional sensitivity are currently unknown, and the role of surround inhibition is unclear. In magnetoencephalogram (MEG), large high-contrast drifting gratings induce a strong gamma response (GR), which also attenuates with an increase in the gratings’ velocity. It has been suggested that the slope of this GR attenuation is mediated by inhibitory interactions in the primary visual cortex. Herein, we investigate whether SS is related to this inhibitory-based MEG measure. We evaluated SS and GR in two independent samples of participants: school-age boys and adult women. The slope of GR attenuation predicted inter-individual differences in SS in both samples. Test-retest reliability of the neuro-behavioral correlation was assessed in the adults, and was high between two sessions separated by several days or weeks. Neither frequencies nor absolute amplitudes of the GRs correlated with SS, which highlights the functional relevance of velocity-related changes in GR magnitude caused by augmentation of incoming input. Our findings provide evidence that links the psychophysical phenomenon of SS to inhibitory-based neural responses in the human primary visual cortex. This supports the role of inhibitory interactions as an important underlying mechanism for spatial suppression. Highlights The role of surround inhibition in perceptual spatial suppression (SS) is debated GR attenuation with increasing grating’s velocity may reflect surround inhibition People with greater GR attenuation exhibit stronger SS The neuro-behavioral correlation is replicated in school-age boys and adult women The surround inhibition in the V1 is an important mechanism underlying SS
... or from physical salience (e.g., a red book among blue books will automatically capture 83 attention) (Beck and Kastner, 2009;Desimone and Duncan, 1995;Egeth and Yantis, 1997). 84 ...
Article
Studies of selective attention typically consider the role of task goals or physical salience, but attention can also be captured by previously reward-associated stimuli, even if they are currently task irrelevant. One theory underlying this value-driven attentional capture (VDAC) is that reward-associated stimulus representations undergo plasticity in sensory cortex, thereby automatically capturing attention during early processing. To test this, we used magnetoencephalography to probe whether stimulus location and identity representations in sensory cortex are modulated by reward learning. We furthermore investigated the time course of these neural effects, and their relationship to behavioral VDAC. Male and female human participants first learned stimulus–reward associations. Next, we measured VDAC in a separate task by presenting these stimuli in the absence of reward contingency and probing their effects on the processing of separate target stimuli presented at different time lags. Using time-resolved multivariate pattern analysis, we found that learned value modulated the spatial selection of previously rewarded stimuli in posterior visual and parietal cortex from;260 ms after stimulus onset. This value modulation was related to the strength of participants’ behavioral VDAC effect and persisted into subsequent target processing. Importantly, learned value did not influence cortical signatures of early processing (i.e., earlier than;200 ms); nor did it influence the decodability of stimulus identity. Our results suggest that VDAC is underpinned by learned value signals that modulate spatial selection throughout posterior visual and parietal cortex. We further suggest that VDAC can occur in the absence of changes in early visual processing in cortex.
... Because the guidance of gaze under natural circumstances is tightly linked to the allocation of spatial attention (e.g., Deubel & Schneider, 1996), it is of interest to note that attention allocation is frequently viewed as a result of competition between items in conjunction with a mechanism that controls priority (Desimone & Duncan, 1995;Schneider, Einhäuser, & Horstmann, 2013). Adopting this view, we can interpret the present results on the allocation of attention and gaze during scene viewing as follows: the currently fixated location competes for attention with potential future locations. ...
Article
Full-text available
Fixation durations provide insights into processing demands. We investigated factors controlling fixation durations during scene viewing in two experiments. In Experiment 1, we tested the degree to which fixation durations adapt to global scene processing difficulty by manipulating the contrast (from original contrast to isoluminant) and saturation (original vs. grayscale) of the entire scene. We observed longer fixation durations for lower levels of contrast, and longer fixation durations for grayscale than for color scenes. Thus fixation durations were globally slowed as visual information became more and more degraded, making scene processing increasingly more difficult. In Experiment 2, we investigated two possible sources for this slow-down. We used "checkerboard" stimuli in which unmodified patches alternated with patches from which luminance information had been removed (isoluminant patches). Fixation durations showed an inverted immediacy effect (longer, rather than shorter, fixation durations on unmodified patches) along with a parafoveal-on-foveal effect (shorter fixation durations, when an unmodified patch was fixated next). This effect was stronger when the currently fixated patch was isoluminant as opposed to unmodified. Our results suggest that peripheral scene information substantially affects fixation durations and are consistent with the notion of competition among the current and potential future fixation locations.
... Managing the limited capacity of working memory (WM) requires control processes that can prioritize goal-relevant information and suppress goal-irrelevant information [1,2]. Prioritization and suppression processes can operate prospectively on perceptual input, with exogenous selective attention serving as an ''input gate'' that controls what perceptual information is ultimately encoded into WM [4][5][6], or retrospectively on the internally maintained contents of WM, effectively serving as an ''output gate'' that controls what mnemonic information is most likely to guide future behavior [7][8][9][10]. For both prospective and retrospective control, substantial evidence has shown that prioritization is associated with theta oscillations that are typically observed over frontal cortex [3,11,12,13], whereas suppression is associated with alpha oscillations that are typically observed over occipital-parietal cortex [3,11,[14][15][16][17]. ...
Article
Working memory (WM) relies on the prioritization of relevant information and suppression of irrelevant information [1, 2]. Prioritizing relevant information has been linked to theta frequency neural oscillations in lateral prefrontal cortex and suppressing irrelevant information has been linked to alpha oscillations in occipito-parietal cortex [3,11]. Here, we used a retrospective-cue WM paradigm to manipulate prioritization and suppression task demands designed to drive theta oscillations in prefrontal cortex and alpha oscillations in parietal cortex, respectively. To causally test the role of these neural oscillations, we applied rhythmic transcranial magnetic stimulation (TMS) in either theta or alpha frequency to prefrontal and parietal regions identified using functional MRI. The effect of rhythmic TMS on WM performance was dependent on whether the TMS frequency matched or mismatched the expected underlying task-driven oscillations of the targeted region. Functional MRI in the targeted regions predicted subsequent TMS effects across subjects supporting a model by which theta oscillations are excitatory to neural activity, and alpha oscillations are inhibitory. Together, these results causally establish dissociable roles for prefrontal theta oscillations and parietal alpha oscillations in the control of internally maintained WM representations.
... Humans are bombarded with visual stimuli throughout their daily routines, with most of this information being irrelevant to current task goals. Due to the limited capacity of cognitive resources (Desimone and Duncan, 1995;Corbetta and Shulman, 2002), we must selectively attend to stimuli most relevant to our ongoing tasks (Yantis, 2000;Theeuwes, 2010) and those that provide critical information pertaining to contingencies within the environment (Miskovic and Keil, 2012). This enhanced attentional processing of important information impacts downstream cognitive systems (Hillyard et al., 1998), such as working memory (Awh et al., 2006;Ikkai and Curtis, 2011;Gazzaley and Nobre, 2012). ...
Article
Full-text available
Previous work suggests that threat-related stimuli are stored to a greater degree in working memory compared to neutral stimuli. However, most of this research has focused on stimuli with physically salient threat attributes (e.g., angry faces), failing to account for how a “neutral” stimulus that has acquired threat-related associations through differential aversive conditioning influences working memory. The current study examined how differentially conditioned safe (i.e., CS–) and threat (i.e., CS+) stimuli are stored in working memory relative to a novel, non-associated (i.e., N) stimuli. Participants (n = 69) completed a differential fear conditioning task followed by a change detection task consisting of three conditions (CS+, CS–, N) across two loads (small, large). Results revealed individuals successfully learned to distinguishing CS+ from CS– conditions during the differential aversive conditioning task. Our working memory outcomes indicated successful load manipulation effects, but no statistically significant differences in accuracy, response time (RT), or Pashler’s K measures of working memory capacity between CS+, CS–, or N conditions. However, we observed significantly reduced RT difference scores for the CS+ compared to CS– condition, indicating greater RT differences between the CS+ and N condition vs. the CS– and N condition. These findings suggest that differentially conditioned stimuli have little impact on behavioral outcomes of working memory compared to novel stimuli that had not been associated with previous safe of aversive outcomes, at least in healthy populations.
... Through interviews with the participants after the experiment, we found that, in VR, except for the region being viewed, the other areas are blurred in the users' field of vision (FOV). The HMD has a dynamic blur-rendering strategy that simulates the selective visual attention of human beings (Desimone and Duncan 1995). Such a technique improves the refreshing rate, but, since the peripheral area is ambiguous, the rate is very much influenced by users' peripheral vision, which has been shown to be important for environmental perception (Rayner 2009;Franchak and Adolph 2010). ...
Article
Maps based on virtual reality (VR) are evolving and are being increasingly used in the field of geography. However, the advantages of VR based on the map use processes of users over desktop-based environments (DEs) are not fully understood. In this study, an experiment was conducted in which 120 participants performed map use tasks using maps and globes in VR and DE. The participants’ eye movements and questionnaires were collected to compare the map use performance differences. We analyzed the general metrics, information searching and processing metrics of participants (e.g. response time, RT; average fixation duration, AFD; average saccade duration, ASD; saccade frequency, SF, etc.) using maps and globes in different environments. We found that the participants using VR processed information more efficiently (AFDDE = 233.34 ms, AFDVR = 173.09 ms), and the participants using DE had both a significantly shorter response time (RTDE = 88.68 s, RTVR = 124.05 s) and a shorter visual search time (ASDDE = 60.78 ms, ASDVR = 112.13 ms; SFDE = 6.30, SFVR = 2.07). We also found similarities in accuracy, satisfaction and readability. These results are helpful for designing VR maps that can adapt to human cognition and reflect the advantages of VR.
... Several researchers have investigated the relationship between an individual's accuracy on cognitive map tasks and their level of familiarity with to-be-recalled layouts (e.g., Lloyd & Patton, 2011). Frequent experience with spatial targets allows people to precisely know where a target landmark is located in map-like tasks, and in perceptual and visual search tasks, and to solve difficult distance knowledge problems (e.g., Desimone & Duncan, 1995). So, familiarity with a physical environment allows people to solve location tasks (de Goede & Postma, 2015). ...
Article
Humans tend to encode the environment by means of two types of spatial relations: coordinate (metric) and categorical (nonmetric). The present research contributes to methods for disentangling the contribution of categorical and coordinates spatial relations in sketch maps and investigates the role of familiarity with spatial information in accurate encoding of these spatial relations. The results of three experiments show that as familiarity with spatial layout increases, differences between categorical and coordinate spatial relations tend to decrease. Moreover, they reveal that the way in which spatial information has been acquired – through navigation or map study - affects performance. Navigation favours coordinate encoding, while map study favours categorical encoding. Finally, gender differences did not emerge when spatial information was acquired through navigation, but they were present in the case of map study acquisition. In conclusion, it seems possible to extract reliable and independent information on both categorical and coordinate spatial mental representations using sketch maps.
... First, few studies have distinguished two important components of attention, namely, goal-directed and stimulusdriven attention (Corbetta and Shulman, 2002). The former is top-down controlled attention (i.e., it focuses on relevant signals derived from task demands; Annic et al., 2016), whereas the latter is bottom-up driven attention (i.e., it is captured by salient properties of stimuli that are usually irrelevant with the task; Desimone and Duncan, 1995). The tasks used in previous studies -for example, the dot probe task and the visual search task -capture both elements of attention. ...
Article
Full-text available
Attentional biases have received considerable focus in research on cognitive biases and body dissatisfaction (BD). However, most work has focused on spatial allocation of attention. The current two experiments employed a rapid serial visual presentation (RSVP) task to investigate attention bias to body-related words in the temporal domain among young females with high and low BD. During this task, there were two targets presented in the same stimulus stream. The first target was defined as target one (T1) and the second was defined as target 2 (T2). Participants were asked to identify T2 while ignoring T1 in single task mode or identify both targets in the dual task mode. In the current study, Experiment 1 assessed the stimulus-driven attention of body-related stimuli. Participants were required to identify a target of neutral word (T2) as quickly and accurately as possible while ignoring the preceding target (T1) of neutral, fat-, or thin-related words. As expected, we observed spontaneous attentional blink (AB) effects elicited by both fat- and thin-related T1s among participants with high BD, suggesting enhanced awareness of body-related stimuli even when this information does not have to be identified. Such effects did not emerge among participants without BD. Experimental 2 investigated the goal-directed attention of body-related stimuli, during which participants needed to identify both the T1 and neutral T2. Participants with BD showed reduced AB effects after both fat- and thin-related T1, suggesting facilitated consolidation of body-related information in goal-directed attention among participants with BD. These findings have important clinical implications that it provided insight for creating more accurate attention bias modification (ABM) task aiming at reducing and preventing BD among young females.
... Visual attention is easily controlled by the "objects" in human brains through top-down feedback, especially in the case of object detection. This is explained by the "directed competition theory" in cognitive science (Beck and Kastner 2009;Desimone 1998;Desimone and Duncan 1995) that the feedback transmits high-level semantic information to low-level perception. In addition to the feedforward process, the selectivity of neural activation is also controlled through additional circulation (Kruger et al. 2013), thereby reducing the opportunity to recognize the disturbed. ...
Article
Full-text available
Visual object recognition is one of the most fundamental and challenging research topics in the field of computer vision. The research on the neural mechanism of the primates’ recognition function may bring revolutionary breakthroughs in brain-inspired vision. This Review aims to systematically review the recent works on the intersection of computational neuroscience and computer vision. It attempts to investigate the current brain-inspired object recognition models and their underlying visual neural mechanism. According to the technical architecture and exploitation methods, we describe the brain-inspired object recognition models and their advantages and disadvantages in realizing brain-inspired object recognition. We focus on analyzing the similarity between the artificial and biological neural network, and studying the biological credibility of the current popular DNN-based visual benchmark models. The analysis provides a guide for researchers to measure the occasion and condition when conducting visual object recognition research.
... It has been suggested that the executive attention component of WMC reflects a domain general attentional resource for actively maintaining goal-relevant neural representations (e.g., motivational context, task rules, stimulus-response associations, object features/ attributes) as an 'attentional set' or 'attentional template' that biases lower-level perceptual, cognitive, and behavioral processes through top-down control (Corbetta & Shulman, 2002;Desimone & Duncan, 1995;Kane & Engle, 2002). This theory provides a parsimonious account of executive attention, as the mechanisms supporting goal-directed cognition and behavior in novel contexts and interference-rich environments, such as those encountered during performance of stimulus-response compatibility and response inhibition tasks, tests of EF, and general intelligence tasks (Brydges, Reid, Fox, & Anderson, 2012;Engel de Abreu, Conway, & Gathercole, 2010;Kane & Engle, 2002;McCabe, Roediger, McDaniel, Balota, & Hambrick, 2010). ...
Article
Executive Function (EF) and Effortful Control (EC) have traditionally been viewed as distinct constructs related to cognition and temperament during development. More recently, EF and EC have both been implicated in top‐down self‐regulation ‐ the goal‐directed control of cognition, emotion, and behavior. We propose that executive attention, a limited‐capacity attentional resource subserving goal‐directed cognition and behavior, is the common cognitive mechanism underlying the self‐regulatory capacities captured by EF and EC. We addressed three related questions: 1) Do behavioral ratings of EF and EC represent the same self‐regulation construct? 2) Is this self‐regulation construct explained by a common executive attention factor as measured by performance on cognitive tasks? and 3) Does the executive attention factor explain additional variance in attention deficit hyperactivity disorder (ADHD) problems to behavioral ratings of self‐regulation? Measures of performance on complex memory span, general intelligence, and response inhibition tasks were obtained from 136 preadolescent children (M = 11 years, 10 months, SD = 8 months), along with self‐ and parent‐reported EC, and parent‐reported EF, and ADHD problems. Results from structural equation modeling demonstrated that behavioral ratings of EF and EC measured the same self‐regulation construct. Cognitive tasks measured a common executive attention factor that significantly explained 30% of the variance in behavioral ratings of self‐regulation. Executive attention failed to significantly explain additional variance in ADHD problems beyond that explained by behavioral ratings of self‐regulation. These findings raise questions about the utility of task‐based cognitive measures in research and clinical assessment of self‐regulation and psychopathology in developmental samples.
... Prefrontal cortical areas encode the valence of stimuli and influence primary sensory cortices through corticocortical projections in a top-down fashion (Bidet-Caulet et al., 2015;Desimone and Duncan, 1995;Miller and Cohen, 2001;Romanski and Goldman-Rakic, 2002;Tomita et al., 1999;Zanto et al., 2011;Zhang et al., 2014). Among these prefrontal areas, the anterior cingulate cortex (ACC) may be an ideal structure for mediating the interaction between an aversive cue and sound-evoked flight behavior. ...
Article
Full-text available
For survival, animals encode prominent events in complex environments, which modulates their defense behavior. Here, we design a paradigm that assesses how a mild aversive cue (i.e., mild air puff) interacts with sound-evoked flight behavior in mice. We find that air puffing facilitates sound-evoked flight behavior by enhancing the auditory responses of auditory cortical neurons. We then find that the anterior part of the anterior cingulate cortex (ACC) encodes the valence of air puffing and modulates the auditory cortex through anatomical examination, physiological recordings, and optogenetic/chemogenetic manipulations. Activating ACC projections to the auditory cortex simulates the facilitating effect of air puffing, whereas inhibiting the ACC or its projections to the auditory cortex neutralizes this facilitating effect. These findings show that the ACC regulates sound-evoked flight behavior by potentiating neuronal responses in the auditory cortex.
... Inspired by research on the human visual system, human neurology [46] and cognitive psychology [12], salient object detection (SOD) has become a topic of interest in the field of computer vision. Specifically, SOD technology simulates how the human visual system detects objects of interest in a scene and provides computer vision with this human-like ability. ...
Article
Full-text available
Most of existing salient object detection models are based on fully convolutional network (FCN), which learn multi-scale/level semantic information through convolutional layers to obtain high-quality predicted saliency maps. However, convolution is locally interactive, it is difficult to capture remote dependencies, and FCN-based methods suffer from coarse object boundaries. In this paper, to solve these problems, we propose a novel transformer framework for salient object detection (named TF-SOD), which mainly consists of the encoder part of the FCN, fusion module (FM), transformer module (TM) and feature decoder module (FDM). Specifically, FM is a bridge connecting the encoder and TM and provides some foresight for the non-local interaction of TM. Besides, FDM can efficiently decode the non-local features output by TM and achieve deep fusion with local features. This architecture enables the network to achieve a close integration of local and non-local interactions, making information complementary to each other, deeply mining the associated information between features. Furthermore, we also propose a novel edge reinforcement learning strategy, which can effectively suppress edge blurring from local and global aspects by means of powerful network architecture. Extensive experiments using five datasets demonstrate that the proposed method outperforms 19 state-of-the-art methods
... They carry out a local analysis of the image elements in their small receptive fields (RFs) and 16 feed the visual information forward to higher visual areas. Neurons in higher areas have larger RFs 17 and integrate information to represent increasingly abstract features of the visual scene, including 18 object category and identity (Desimone and Duncan, 1995;Yamins and DiCarlo, 2016). However, there 19 are many images for which the analysis is not complete when information has reached the higher 20 visual areas ( Kar et al., 2019). ...
... During search for a visual object, a mental representation of the target object is maintained in visual working memory to guide attention toward potentially task-relevant regions (Desimone and Duncan, 1995;Olivers and Eimer, 2011). In everyday situations, individuals may oftentimes try to find multiple objects at the same time, which would require the maintenance of more than one target representation. ...
Article
Full-text available
Cognitive control can involve proactive (preparatory) and reactive (corrective) mechanisms. Using a gaze-contingent eye tracking paradigm combined with fMRI, we investigated the involvement of these different modes of control and their underlying neural networks, when switching between different targets in multiple-target search. Participants simultaneously searched for two possible targets presented among distractors, and selected one of them. In one condition, only one of the targets was available in each display, so that the choice was imposed, and reactive control would be required. In the other condition, both targets were present, giving observers free choice over target selection, and allowing for proactive control. Switch costs emerged only when targets were imposed and not when target selection was free. We found differential levels of activity in the frontoparietal control network depending on whether target switches were free or imposed. Furthermore, we observed core regions of the default mode network to be active during target repetitions, indicating reduced control on these trials. Free and imposed switches jointly activated parietal and posterior frontal cortices, while free switches additionally activated anterior frontal cortices. These findings highlight unique contributions of proactive and reactive control during visual search.
... Also, related work in the literature usually focuses on sensory perception only, and does not normally involve flexible switching between sensory domains, contrary to the paradigm considered here. Indeed, a similar change in domain selectivity was found in studies where attentional state was directed either at location or a specific feature dimension (Bichot et al., 2015;Desimone and Duncan, 1995;Liu et al., 2007). This led to differences in behaviour, improving performance related either to spatial or feature attention respectively. ...
Article
Full-text available
Many recent advances in artificial intelligence (AI) are rooted in visual neuroscience. However, ideas from more complicated paradigms like decision-making are less used. Although automated decision-making systems are ubiquitous (driverless cars, pilot support systems, medical diagnosis algorithms etc.), achieving human-level performance in decision making tasks is still a challenge. At the same time, these tasks that are hard for AI are easy for humans. Thus, understanding human brain dynamics during these decision-making tasks and modeling them using deep neural networks could improve AI performance. Here we modelled some of the complex neural interactions during a sensorimotor decision making task. We investigated how brain dynamics flexibly represented and distinguished between sensory processing and categorization in two sensory domains: motion direction and color. We used two different approaches for understanding neural representations. We compared brain responses to 1) the geometry of a sensory or category domain (domain selectivity) and 2) predictions from deep neural networks (computation selectivity). Both approaches gave us similar results. This confirmed the validity of our analyses. Using the first approach, we found that neural representations changed depending on context. We then trained deep recurrent neural networks to perform the same tasks as the animals. Using the second approach, we found that computations in different brain areas also changed flexibly depending on context. Color computations appeared to rely more on sensory processing, while motion computations more on abstract categories. Overall, our results shed light to the biological basis of categorization and differences in selectivity and computations in different brain areas. They also suggest a way for studying sensory and categorical representations in the brain: compare brain responses to both a behavioral model and a deep neural network and test if they give similar results.
... Visual attention dramatically affects perception and a wide variety of measures of neural activity in essentially every visual and visuomotor brain area (for reviews, see Maunsell 1 and Moore and Zirnsak 2 ). Attention flexibly modulates signatures of neuronal activity, including trial-averaged firing rates, 1,3,4 shared variability between pairs of neurons in the same [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] and different brain areas, 15,[21][22][23] interdependence of neuronal populations on a range of timescales, 15,[24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43] and the dimensionality of population activity within each brain area. [44][45][46] The behavioral effects of attention make it clear that visual information can be flexibly routed: a stimulus can either guide or be unrelated to a perceptual decision, depending on the task condition. ...
Article
Visual attention allows observers to change the influence of different parts of a visual scene on their behavior, suggesting that information can be flexibly shared between visual cortex and neurons involved in decision making. We investigated the neural substrate of flexible information routing by analyzing the activity of populations of visual neurons in the medial temporal area (MT) and oculo-motor neurons in the superior colliculus (SC) while rhesus monkeys switched spatial attention. We demonstrated that attention increases the efficacy of visuomotor communication: trial-to-trial variability in SC population activity could be better predicted by the activity of the MT population (and vice versa) when attention was directed toward their joint receptive fields. Surprisingly, this improvement in prediction was not explained by changes in the dimensionality of the shared subspace or in the magnitude of local or shared pairwise noise correlations. These results lay a foundation for future theoretical and experimental studies into how visual attention can flexibly change information flow between sensory and decision neurons.
... Participants in both groups increased gaze dwell times towards both of these regions of interest while the agent approached, and did so more strongly with regard to gaze on objects (especially in the PC group). The higher amount of gaze allocated towards the objectsespecially after the first trials, when gaze towards the head was still more frequentis most parsimoniously explained by a novelty effect (Desimone, 1995): while participants saw the same character throughout the experiment, the object changed every trial. The increase in gaze towards either the character's head or the object as it approached may be due to a better visibility to the participants as well as to increased measurement accuracy, and this effect may interact with the generally larger interest in gazing at the objects. ...
Article
Full-text available
People show a robust tendency to gaze at other human beings when viewing images or videos, but were also found to relatively avoid gaze at others in several real‐world situations. This discrepancy, along with theoretical considerations, spawned doubts about the appropriateness of classical laboratory‐based experimental paradigms in social attention research. Several researchers instead suggested the use of immersive virtual scenarios in eliciting and measuring naturalistic attentional patterns, but the field, struggling with methodological challenges, still needs to establish the advantages of this approach. Here, we show using eye‐tracking in a complex social scenario displayed in virtual reality that participants show enhanced attention towards the face of an avatar at near distance and demonstrate an increased reactivity towards her social gaze as compared to participants who viewed the same scene on a computer monitor. The present study suggests that reactive virtual agents observed in immersive virtual reality can elicit natural modes of information processing and can help to conduct ecologically more valid experiments while maintaining high experimental control.
... An influential idea is that behaviorally relevant visual input can be favored over irrelevant visual input by preparatory activation of neural populations that represent visual properties of the relevant (i.e., target) object, such as its color, shape, or size. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15] These visual properties, however, are unknown during real-world search, because they depend on the (inherently unknown) location of the target object. The retinal image size of an object, for instance, depends on the distance between observer and object. ...
Article
Humans are remarkably proficient at finding objects within complex visual scenes. According to current theories of attention,1, 2, 3 visual processing of an object of interest is favored through the preparatory activation of object-specific representations in visual cortex.4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 One key problem that is inherent to real-world visual search but is not accounted for by current theories is that a given object will produce a dramatically different retinal image depending on its location, which is unknown in advance. For instance, the color of the retinal image depends on the illumination on the object, its shape depends on the viewpoint, and (most critically) its size can vary by several orders of magnitude, depending on the distance to the observer. In order to benefit search, preparatory activity thus needs to incorporate contextual expectations. In the current study, we measured fMRI blood-oxygen-level-dependent (BOLD) activity in human observers while they prepared to search for objects at different distances in indoor-scene photographs. First, we established that observers instantiated preparatory object representations: activity patterns in object-selective cortex evoked during search preparation (while no objects were presented) resembled activity patterns evoked by viewing those objects in isolation. Second, we demonstrated that these preparatory object representations were systematically modulated by expectations derived from scene context: activity patterns reflected the predicted retinal image of the object at each distance (i.e., distant search evoking smaller object representations and nearby search evoking larger object representations). These findings reconcile current theories of attentional selection with the challenges of real-world vision.
... The flexible attention system can choose the split or unitary mode according to the task requirement (Jefferies & Di Lollo, 2009;Jefferies et al., 2014). Some researchers hold that unitary attention is the default option of the attentional system while split attention is a special case, and these views are based on the neuronal substrate of spatial attention (Koch & Ullman, 1985;Kusunoki et al., 2000) or the biased competition model (Cave et al., 2010;Desimone & Duncan, 1995). These researchers believe that the existence of split attention cannot easily be reconciled with the presented theories or models (Jans et al., 2010). ...
Article
Full-text available
Although it is often assumed that spatial attention exists in the form of a unitary focus, the split-attention hypothesis proposes that attention can be simultaneously divided into two spatially noncontiguous positions and that the space in between can be ignored. However, whether split attention occurs directly based on the generation of attentional benefit or whether it requires a gradual divide from a unitary focus over time has not been clarified. In the present study, by using two spatial salient cues to direct the attention allocation of participants, we aimed to investigate whether attention requires time to divide from a unitary focus and whether the appearance time of split attention varies when the task difficulty level increases between experiments. The results showed that attention required time to divide from a unitary focus, and the position between the two cued positions was not excluded by attention when the stimulus-onset asynchrony (SOA) was 60 ms. However, as the task difficulty increased between experiments, the appearance time of split attention was earlier. These findings suggest that the appearance time of split attention has a certain flexibility and can be changed according to the task requirement, thus implying that split attention and unitary attention present some common attention mechanisms and that a split or unitary mode can be flexibly selected for an attention system.
... In contrast, feature-based attention focuses on the selection of attended visual features (such as upward motion or the color red) irrespective of stimulus location (Liu, 2019;Maunsell and Treue, 2006). Selective attention generally enhances (or gains) the responses of sensory neurons to stimuli/features in the focus of attention and diminishes responses to distractors, an effect termed ''gain modulation'' (Buschman and Kastner, 2015;Desimone and Duncan, 1995;Maunsell and Treue, 2006;Squire et al., 2013). However, the origins of these neuronal correlates of visual selective attention in sensory areas have been implicated to lie in parietal (Saalmann et al., 2007) and especially frontal cortical regions that exert top-down modulatory influence (Buschman and Miller, 2007;Everling et al., 2002;Rainer et al., 1998a). ...
Article
Full-text available
Feature-based attention enables privileged processing of specific visual properties. During feature-based attention, neurons in visual cortices show “gain modulation” by enhancing neuronal responses to the features of attended stimuli due to top-down signals originating from prefrontal cortex (PFC). Attentional modulation in visual cortices requires “feature similarity:” neurons only increase their responses when the attended feature variable and the neurons’ preferred feature coincide. However, whether gain modulation based on feature similarity is a general attentional mechanism is currently unknown. To address this issue, we record single-unit activity from PFC of macaques trained to switch attention between two conjunctive feature parameters. We find that PFC neurons experience gain modulation in response to attentional demands. However, this attentional gain modulation in PFC is independent of the feature-tuning preferences of neurons. These findings suggest that feature similarity is not a general mechanism in feature-based attention throughout the cortical processing hierarchy.
... These internally directed cognitions can be either goal-directed, such as in deliberate planning (Spreng et al., 2010), imagination (Benedek, 2018;Zabelina, 2018) and problem-solving (Jung-Beeman et al., 2004;Kounios & Beeman, 2009;Salvi et al., 2015), or spontaneous, such as during mind wandering (Smallwood & Schooler, 2006). As we are limited in our information processing ability (Desimone & Duncan, 1995), internally directed cognition (IDC) and externally directed cognition (EDC) are considered competing states (Chun et al., 2011), sharing a common pool of cognitive resources (Verschooren et al., 2019). Yet, IDC and EDC rely on different neural mechanisms (Dixon et al., 2014) and thus are characterized by distinct neurophysiological signatures as evidenced by research using fMRI Dixon et al., 2018;Kucyi et al., 2017;Margulies et al., 2016;C. ...
Article
Full-text available
Many goal-directed, as well as spontaneous everyday activities (e.g., planning, mind-wandering), rely on an internal focus of attention. In this fMRI–eye-tracking coregistration study, we investigated brain mechanisms and eye behavior related to internally versus externally directed cognition. Building on an established paradigm, we manipulated internal attention demands within tasks utilizing conditional stimulus masking. Internally directed cognition involved bilateral activation of the lingual gyrus and inferior parietal lobe areas as well as wide-spread deactivation of visual networks. Moreover, internally directed cognition was related to greater pupil diameter, pupil diameter variance, blink duration, fixation disparity variance, and smaller amounts of microsaccades. FMRI–eye-tracking covariation analyses further revealed that larger pupil diameter was related to increased activation of basal ganglia and lingual gyrus. It can be concluded that internally and externally directed cognition are characterized by distinct neurophysiological signatures. The observed neurophysiological differences indicate that internally directed cognition is associated with reduced processing of task-irrelevant information and increased mental load. These findings shed further light on the interplay between neural and perceptual mechanisms contributing to an internal focus of attention.
... Visual attention dramatically affects perception and a wide variety of measures of neural activity 18 in essentially every visual and visuomotor brain area (for reviews, see (Maunsell, 2015;Moore 19 and Zirnsak, 2017)). Attention flexibly modulates signatures of neuronal activity including trial-20 averaged firing rates (Desimone and Duncan, 1995 reach asymptote) predicted SC activity at least as well as a full linear model (fit using ridge 175 regression; see Methods). The prediction accuracy for the attend in trials was significantly better 176 than the attend out trials irrespective of the number of predictive dimensions (the black line is 177 always above the gray line in Figure 2f). ...
Preprint
Full-text available
Visual attention allows observers to flexibly use or ignore visual information, suggesting that information can be flexibly routed between visual cortex and neurons involved in decision-making. We investigated the neural substrate of flexible information routing by analyzing the activity of populations of visual neurons in the medial temporal area (MT) and oculomotor neurons in the superior colliculus (SC) while rhesus monkeys switched spatial attention. We demonstrated that attention increases the efficacy of visuomotor communication: trial-to-trial variability of the population of SC neurons was better predicted by the activity of MT neurons (and vice versa) when attention was directed toward their joint receptive fields. Surprisingly, this improvement in prediction was not explained or accompanied by changes in the dimensionality of the shared subspace or in local or shared pairwise noise correlations. These results suggest a mechanism by which visual attention can affect perceptual decision-making without altering local neuronal representations.
... This finding is consistent with previous studies showing that disruption of holistic processing impairs their detection during visual search (Hershler & Hochstein, 2005;Lewis & Edmonds, 2003;Rousselet, Macé, & Fabre-Thorpe, 2003). In the context of visual search, it has been proposed that when searching for a target among distractors, a "search template" of target information would be used to prioritize and facilitate the processing of information matching this template in order to increase search efficiency (Desimone & Duncan, 1995). Under this assumption, the lower accuracy observed to detect face targets when holistic processing is disrupted would indicate that holistic representations are used as search template for faces. ...
Article
Full-text available
Previous studies have shown that face stimuli influence the programming of eye movements by eliciting involuntary and extremely fast saccades toward them. The present study examined whether holistic processing of faces mediates these effects. We used a saccadic choice task in which participants were presented simultaneously with two images and had to perform a saccade toward the one containing a target stimulus (e.g., a face). Across three experiments, stimuli were altered via upside-down inversion (Experiment 1) or scrambling of thumbnails within the images (Experiments 2 and 3) in order to disrupt holistic processing. We found that disruption of holistic processing only had a limited impact on the latency of saccades toward face targets, which remained extremely short (minimum saccadic reaction times of only ∼120-130 ms), and did not affect the proportion of error saccades toward face distractors that captured attention more than other distractor categories. It, however, resulted in increasing error rate of saccades toward face targets. These results suggest that the processing of isolated face features is sufficient to elicit extremely fast and involuntary saccadic responses toward them. Holistic representations of faces may, however, be used as a search template to accurately detect faces.
... Different machine-learning-based normalization methods are distinguished based on their choice of a normalization pool. Attention Attention has been extensively studied in neuroscience (Desimone and Duncan, 1995;Carrasco, 2011). Computational models are able to capture various aspects of bottom-up (Koch and Ullman, 1987) and top-down attention (Reynolds and Heeger, 2009). ...
Article
Artificial neural networks (ANNs) are essential tools in machine learning that have drawn increasing attention in neuroscience. Besides offering powerful techniques for data analysis, ANNs provide a new approach for neuroscientists to build models for complex behaviors, heterogeneous neural activity, and circuit connectivity, as well as to explore optimization in neural systems, in ways that traditional models are not designed for. In this pedagogical Primer, we introduce ANNs and demonstrate how they have been fruitfully deployed to study neuroscientific questions. We first discuss basic concepts and methods of ANNs. Then, with a focus on bringing this mathematical framework closer to neurobiology, we detail how to customize the analysis, structure, and learning of ANNs to better address a wide range of challenges in brain research. To help readers garner hands-on experience, this Primer is accompanied with tutorial-style code in PyTorch and Jupyter Notebook, covering major topics.
... Goal-driven selective attention has provided a successful paradigm for investigating the sources and mechanisms of top-down modulation of signal processing within perceptual streams. Decades of research have yielded enormous progress in revealing how the locations and feature-related attributes of relevant events are prioritised and integrated along the sensory hierarchies (Desimone & Ducan 1995;Fries 2015;Kastner & Ungerleider 2000;Reynolds & Chelazzi 2004). These top-down biases were subsequently shown also to carry dynamic information about the estimated timing of relevant events -a phenomenon called temporal orienting of attention or, more generally, temporal expectation (Nobre & Rohenkohl 2014). ...
Preprint
Full-text available
The fundamental role that our long-term memories play in guiding perception is increasingly recognised, but the functional and neural mechanisms are just beginning to be explored. Though experimental approaches are being developed to investigate the influence of long-term memories on perception, these remain mostly static and neglect their temporal and dynamic nature. Here we show we show that our long-term memories can guide attention proactively and dynamically based on learned temporal associations. Across two experiments we found that detection and discrimination of targets appearing within previously learned contexts are enhanced when the timing of target appearance matches the learned temporal contingency. Neural markers of temporal preparation revealed that the learned temporal associations trigger specific temporal predictions. Our findings emphasize the ecological role that memories play in predicting and preparing perception of anticipated events, calling for revision of the usual conceptualisation of contextual associative memory as a reflective and retroactive function.
... Survival depends on successfully navigating a complex, dynamic environment, but the brain has limited processing resources for sampling environmental information (Desimone and Duncan, 1995;Kastner and Ungerleider, 2000). Selective attention comprises a set of neural mechanisms through which some aspects of the environment receive preferential processing relative to others. ...
Article
There has been little evidence linking changes in spiking activity that occur prior to a spatially predictable target (i.e., prior to target selection) to behavioral outcomes, despite such preparatory changes being widely assumed to enhance the sensitivity of sensory processing. We simultaneously recorded from frontal and parietal nodes of the attention network while macaques performed a spatial cueing task. When anticipating a spatially predictable target, different patterns of coupling between spike timing and the oscillatory phase in local field potentials—but not changes in spike rate—were predictive of different behavioral outcomes. These behaviorally relevant differences in local and between-region synchronization occurred among specific cell types that were defined based on their sensory and motor properties, providing insight into the mechanisms underlying enhanced sensory processing prior to target selection. We propose that these changes in neural synchronization reflect differential anticipatory engagement of the network nodes and functional units that shape attention-related sampling.
... These internally directed cognitions can be either goal-directed, such as in deliberate planning (Spreng et al., 2010), imagination (Benedek, 2018;Zabelina, 2018) and problem-solving (Jung-Beeman et al., 2004;Kounios & Beeman, 2009;Salvi et al., 2015), or spontaneous, such as during mind wandering (Smallwood & Schooler, 2006). As we are limited in our information processing ability (Desimone & Duncan, 1995), internally directed cognition (IDC) and externally directed cognition (EDC) are considered competing states (Chun et al., 2011), sharing a common pool of cognitive resources (Verschooren et al., 2019). Yet, IDC and EDC rely on different neural mechanisms (Dixon et al., 2014) and thus are characterized by distinct neurophysiological signatures as evidenced by research using fMRI Dixon et al., 2018;Kucyi et al., 2017;Margulies et al., 2016;C. ...
Preprint
Full-text available
Many goal-directed, as well as spontaneous everyday activities (e.g., planning, mind-wandering), rely on an internal focus of attention. In this fMRI-eye-tracking coregistration study, we investigated brain mechanisms and eye behavior related to internally versus externally directed cognition. Building on an established paradigm, we manipulated internal attention demands within tasks utilizing conditional stimulus masking. Internally directed cognition involved bilateral activation of the lingual gyrus and inferior parietal lobe areas as well as widespread deactivation of visual networks. Moreover, internally directed cognition was related to greater pupil diameter, pupil diameter variance, blink duration, fixation disparity variance, and smaller amounts of microsaccades. FMRI-eye-tracking covariation analyses further revealed that larger pupil diameter was related to increased activation of basal ganglia and lingual gyrus. It can be concluded that internally and externally directed cognition are characterized by distinct neurophysiological signatures. The observed neurophysiological differences indicate that internally directed cognition is associated with reduced processing of task-irrelevant information and increased mental load. These findings shed further light on the interplay between neural and perceptual mechanisms contributing to an internal focus of attention.
... Object detection and localization is also performed in the visual cortex in the ventral and dorsal stream (Desimone and Duncan, 1995). Artificial neural networks like CNNs have taken inspiration from that and are now highly-performant for this task (Ebrahimpour et al., 2019). ...
Article
Full-text available
Frequency-modulated continuous wave radar sensors play an essential role for assisted and autonomous driving as they are robust under all weather and light conditions. However, the rising number of transmitters and receivers for obtaining a higher angular resolution increases the cost for digital signal processing. One promising approach for energy-efficient signal processing is the usage of brain-inspired spiking neural networks (SNNs) implemented on neuromorphic hardware. In this article we perform a step-by-step analysis of automotive radar processing and argue how spiking neural networks could replace or complement the conventional processing. We provide SNN examples for two processing steps and evaluate their accuracy and computational efficiency. For radar target detection, an SNN with temporal coding is competitive to the conventional approach at a low compute overhead. Instead, our SNN for target classification achieves an accuracy close to a reference artificial neural network while requiring 200 times less operations. Finally, we discuss the specific requirements and challenges for SNN-based radar processing on neuromorphic hardware. This study proves the general applicability of SNNs for automotive radar processing and sustains the prospect of energy-efficient realizations in automated vehicles.
... Both learning and attention rely, to varying degrees, on the integration of top-down inputs with bottom-up signals. During attention, higher-order brain regions are thought to provide feedback signals to bias bottom-up information processing (Desimone and Duncan, 1995;Gilbert and Li, 2013), most prominently through direct feedback projections (Leinweber et al., 2017;Zhang et al., 2014) or through thalamic nuclei (Chalupa et al., 1976;Wimmer et al., 2015). These feedback projections can target excitatory or specific inhibitory interneurons (Leinweber et al., 2017;Zhang et al., 2014Zhang et al., , 2016. ...
Article
Full-text available
Selectivity of cortical neurons for sensory stimuli can increase across days as animals learn their behavioral relevance and across seconds when animals switch attention. While both phenomena occur in the same circuit, it is unknown whether they rely on similar mechanisms. We imaged primary visual cortex as mice learned a visual discrimination task and subsequently performed an attention switching task. Selectivity changes due to learning and attention were uncorrelated in individual neurons. Selectivity increases after learning mainly arose from selective suppression of responses to one of the stimuli but from selective enhancement and suppression during attention. Learning and attention differentially affected interactions between excitatory and PV, SOM, and VIP inhibitory cells. Circuit modeling revealed that cell class-specific top-down inputs best explained attentional modulation, while reorganization of local functional connectivity accounted for learning-related changes. Thus, distinct mechanisms underlie increased discriminability of relevant sensory stimuli across longer and shorter timescales.
Preprint
Full-text available
Decision-makers are often faced with limited information about the outcomes of their choices. Current formalizations of uncertain choice, such as the explore-exploit dilemma, do not apply well to decisions in which search capacity can be allocated to each option in variable amounts. Such choices confront decision-makers with the need to tradeoff between breadth - allocating a small amount of capacity to each of many options – and depth - focusing capacity on a few options. We formalize the breadth-depth dilemma through a finite sample capacity model. We find that, if capacity is smaller than 4-7 samples, it is optimal to draw one sample per alternative, favoring breadth. However, for larger capacities, a sharp transition is observed, and it becomes best to deeply sample a very small fraction of alternatives, that decreases with the square root of capacity. Thus, ignoring most options, even when capacity is large enough to shallowly sample all of them, reflects a signature of optimal behavior. Our results also provide a rich casuistic for metareasoning in multi-alternative decisions with bounded capacity.
Preprint
Full-text available
The ability to successfully regulate negative emotions such as fear and anxiety is vital for mental health. The neuropeptide oxytocin (OXT) acts as important modulator of emotion regulation, as reflected by reduced amygdala responses but increased amygdala prefrontal cortex (PFC) functional connectivity in response to threatening stimuli. The present randomized, between-subject, placebo (PLC) controlled pharmacological study combined intranasal administration of OXT with functional MRI during an explicit (cognitive) emotion regulation (i.e. distancing reappraisal) paradigm in 65 healthy male participants to investigate the modulatory effects of OXT on both bottom-up and top-down emotion regulation. OXT attenuated the activation in posterior insular cortex and amygdala during anticipation of top-down regulation of predictable threat stimuli in participants with high trait anxiety, providing evidence to support the anxiolytic action of OXT. In contrast, OXT enhanced amygdala activity during bottom-up anticipation of an unpredictable threat stimulus in participants with low trait anxiety. OXT may thus facilitate top-down goal-directed attention by attenuating amygdala activity in high anxiety individuals, while promote bottom-up attention/vigilance to unexpected threat by enhancing anticipatory amygdala activity in low anxiety individuals. The opposite effects of OXT on anticipatory amygdala activation in high versus low anxiety individuals may suggest a baseline anxiety level dependent mechanism via which OXT promotes optimal levels of amygdala activation during the anticipation of an imminent threat. OXT may thus have the potential to promote an adaptive balance between bottom-up and top-down attention systems depending on individual levels of pre-treatment trait anxiety levels.
Article
Full-text available
The objective of this study is to investigate and to simulate the gaze deployment of observers on paintings. For that purpose, we built a large eye tracking dataset composed of 150 paintings belonging to 5 art movements. We observed that the gaze deployment over the proposed paintings was very similar to the gaze deployment over natural scenes. Therefore, we evaluate existing saliency models and propose a new one which significantly outperforms the most recent deep-based saliency models. Thanks to this new saliency model, we can predict very accurately what are the salient areas of a painting. This opens new avenues for many image-based applications such as animation of paintings or transformation of a still painting into a video clip.
Article
Reward-associated objects receive preferential attention, reflecting a bias in information processing that develops automatically following associative learning. Mounting evidence suggests that such value-driven attention operates abnormally in certain psychopathologies, with attentional biases for reward-associated objects being either exaggerated or blunted compared to healthy controls. Here, I review the evidence linking value-driven attention to psychopathology, including drug addiction, depression, attention-deficit hyperactivity disorder (ADHD), compulsivity, and impulsive and risky decision-making. I conclude by offering an integrative framework for conceptualizing the link between value-driven attention and psychopathology, along with suggestions for future research into this burgeoning area of investigation, including research on object attachment.
Article
Full-text available
The expectation-suppression effect reduced stimulus-evoked responses to expected stimuli – is widely considered to be an empirical hallmark of reduced prediction errors in the framework of predictive coding. Here we challenge this notion by proposing that that expectation suppression could be explained by a reduced attention effect. Specifically, we argue that reduced responses to predictable stimuli can also be explained by a reduced saliency-driven allocation of attention. We base our discussion mainly on findings in the visual cortex and propose that resolving this controversy requires the assessment of qualitative differences between the ways in which attention and ‘surprise’ enhance brain responses.
Preprint
This is a preprint for the manuscript "Monetary and non-monetary reward reduces attentional capture by emotional distractors" (currently under review).
Article
Full-text available
The human brain recurrently prioritizes task-relevant over task-irrelevant visual information. A central, question is whether multiple objects can be prioritized simultaneously. To answer this, we let observers search for two colored targets among distractors. Crucially, we independently varied the number of target colors that observers anticipated, and the number of target colors actually used to distinguish the targets in the display. This enabled us to dissociate the preparation of selection mechanisms from the actual engagement of such mechanisms. Multivariate classification of electroencephalographic activity allowed us to track selection of each target separately across time. The results revealed only small neural and behavioral costs associated with preparing for selecting two objects, but substantial costs when engaging in selection. Further analyses suggest this cost is the consequence of neural competition resulting in limited parallel processing, rather than a serial bottleneck. The findings bridge diverging theoretical perspectives on capacity limitations of feature-based attention.
Article
Full-text available
Attentional modulation of cortical networks is critical for the cognitive flexibility required to process complex scenes. Current theoretical frameworks for attention are based almost exclusively on studies in visual cortex, where attentional effects are typically modest and excitatory. In contrast, attentional effects in auditory cortex can be large and suppressive. A theoretical framework for explaining attentional effects in auditory cortex is lacking, preventing a broader understanding of cortical mechanisms underlying attention. Here, we present a cortical network model of attention in primary auditory cortex (A1). A key mechanism in our network is attentional inhibitory modulation (AIM) of cortical inhibitory neurons. In this mechanism, top-down inhibitory neurons disinhibit bottom-up cortical circuits, a prominent circuit motif observed in sensory cortex. Our results reveal that the same underlying mechanisms in the AIM network can explain diverse attentional effects on both spatial and frequency tuning in A1. We find that a dominant effect of disinhibition on cortical tuning is suppressive, consistent with experimental observations. Functionally, the AIM network may play a key role in solving the cocktail party problem. We demonstrate how attention can guide the AIM network to monitor an acoustic scene, select a specific target, or switch to a different target, providing flexible outputs for solving the cocktail party problem.
Preprint
Identical stimuli can be perceived or go unnoticed across successive presentations, producing divergent behavioral readouts despite similarities in sensory input. We hypothesized that fluctuations in neurophysiological states in the sensory neocortex, which could alter cortical processing at the level of neural subpopulations, underlies this perceptual variability. We analyzed cortical layer-specific electrophysiological activity in visual area V4 during a cued attention task. We find that hit trials are characterized by a larger pupil diameter and lower incidence of microsaccades, indicative of a behavioral state with increased arousal and perceptual stability. Target stimuli presented at perceptual threshold evoke elevated multi-unit activity in V4 neurons in hit trials compared to miss trials, across all cortical layers. Putative excitatory and inhibitory neurons are strongly positively modulated in the input (IV) and deep (V & VI) layers of the cortex during hit trials. Excitatory neurons in the superficial cortical layers exhibit lower variability in hit trials. Deep layer neurons are less phase-locked to low frequency rhythms in hits. Hits are also characterized by greater interlaminar coherence between the superficial and deep layers in the pre-stimulus period, and a complementary pattern between the input layer and both the superficial and deep layers in the stimulus-evoked period. Taken together, these results indicate that a state of elevated levels of arousal and perceptual stability allow enhanced processing of sensory stimuli, which contributes to hits at perceptual threshold.
ResearchGate has not been able to resolve any references for this publication.