Parallel Visual Processing Research Articles

Centralised visual processing center for remote sensing target detection

Target detection in satellite images is an essential topic in the field of remote sensing and computer vision. Despite extensive research efforts, accurate and efficient target detection in remote sensing images remains unsolved due to the large target scale span, dense distribution, and overhead imaging and complex backgrounds, which result in high target feature similarity and serious occlusion. In order to address the above issues in a comprehensive manner, within this paper, we first propose a Centralised Visual Processing Center (CVPC), this structure is a parallel visual processing center for Transformer encoder and CNN, employing a lightweight encoder to capture broad, long-range interdependencies. Pixel-level Learning Center (PLC) module is used to establish pixel-level correlations and improve the depiction of detailed features. CVPC effectively improves the detection efficiency of remote sensing targets with high feature similarity and severe occlusion. Secondly, we propose a centralised feature cross-layer fusion pyramid structure to fuse the results with the CVPC in a top-down manner to enhance the detailed feature representation capability at each layer. Ultimately, we present a Context Enhanced Adaptive Sparse Convolutional Network (CEASC), which improves the accuracy while ensuring the detection efficiency. Based on the above modules, we designed and conducted a series of experiments. These experiments are conducted on three challenging public datasets, DOTA-v1.0, DIOR, and RSDO, showing that our proposed 3CNet achieves a more advanced detection accuracy while balancing the detection speed (78.62% mAP for DOTA-v1.0, 79.12% mAP for DIOR, and 95.50% mAP for RSOD).

Fly eyes are not still: a motion illusion in Drosophila flight supports parallel visual processing.

Most animals shift gaze by a 'fixate and saccade' strategy, where the fixation phase stabilizes background motion. A logical prerequisite for robust detection and tracking of moving foreground objects, therefore, is to suppress the perception of background motion. In a virtual reality magnetic tether system enabling free yaw movement, Drosophila implemented a fixate and saccade strategy in the presence of a static panorama. When the spatial wavelength of a vertical grating was below the Nyquist wavelength of the compound eyes, flies drifted continuously and gaze could not be maintained at a single location. Because the drift occurs from a motionless stimulus - thus any perceived motion stimuli are generated by the fly itself - it is illusory, driven by perceptual aliasing. Notably, the drift speed was significantly faster than under a uniform panorama, suggesting perceptual enhancement as a result of aliasing. Under the same visual conditions in a rigid-tether paradigm, wing steering responses to the unresolvable static panorama were not distinguishable from those to a resolvable static pattern, suggesting visual aliasing is induced by ego motion. We hypothesized that obstructing the control of gaze fixation also disrupts detection and tracking of objects. Using the illusory motion stimulus, we show that magnetically tethered Drosophila track objects robustly in flight even when gaze is not fixated as flies continuously drift. Taken together, our study provides further support for parallel visual motion processing and reveals the critical influence of body motion on visuomotor processing. Motion illusions can reveal important shared principles of information processing across taxa.

The Psychophysical Assessment of Hierarchical Magno-, Parvo- and Konio-Cellular Visual Stream Dysregulations in Migraineurs.

IntroductionAlthough conscious, image-forming illusions have been noted in migraine, few studies have specifically sought to collectively evaluate the role of all three parallel visual processing streams in the retinogeniculostriate pathway involved with image-forming vision and their implications in the development of migraine symptoms.MethodsWe psychophysically assessed the functionality of the inferred magnocellular (MC), parvocellular (PC), and koniocellular (KC) streams at different hierarchical loci across three clinical groups: individuals who experience migraine with aura (MA; n=13), experience migraine without aura (MWO; n=14), and Controls (n=15). Participants completed four experiments: Experiment 1 designed to assess retinal short-wavelength-sensitive (S-) cone sensitivities; Experiment 2 intended to measure postretinal temporal and spatiochromatic contrast sensitivities; Experiment 3 intended to assess postretinal spatiotemporal achromatic contrast sensitivities; and Experiment 4 designed to measure thalamocortical color discriminations along the three cone-excitation axes.ResultsS-cone deficits were revealed with greater retinal areas being affected in MA compared to MWO participants. Findings across the four experiments suggest a prominent retinal locus of dysfunction in MA (lesser in MWO) with potential feedforward compensations occurring within the KC visual stream.ConclusionComplex, integrative network compensations need to be factored in when considering the dysregulating influences of migraine along the visual pathway.

Independent and parallel visual processing of mean, variance, and numerosity: Evidence from dual tasks

Independent and parallel visual processing of mean, variance, and numerosity: Evidence from dual tasks

Independent and parallel visual processing of ensemble statistics: Evidence from dual tasks.

The visual system can represent multiple objects in a compressed form of ensemble summary statistics (such as object numerosity, mean, and feature variance/range). Yet the relationships between the different types of visual statistics remain relatively unclear. Here, we tested whether two summaries (mean and numerosity, or mean and range) are calculated independently from each other and in parallel. Our participants performed dual tasks requiring a report about two summaries in each trial, and single tasks requiring a report about one of the summaries. We estimated trial-by-trial correlations between the precision of reports as well as correlations across observers. Both analyses showed the absence of correlations between different types of ensemble statistics, suggesting their independence. We also found no decrement (except that related to the order of report explained by memory retrieval) in performance in dual compared to single tasks, which suggests that two statistics of one ensemble can be processed in parallel.

Parallel visual processing: its basis and application for the neurological disorders

Parallel visual processing: its basis and application for the neurological disorders

Parallel visual processing and its use for electrodiagnosis of the neuropsychiatric diseases

Parallel visual processing and its use for electrodiagnosis of the neuropsychiatric diseases

Object features and T4/T5 motion detectors modulate the dynamics of bar tracking by Drosophila.

Visual objects can be discriminated by static spatial features such as luminance or dynamic features such as relative movement. Flies track a solid dark vertical bar moving on a bright background, a behavioral reaction so strong that for a rigidly tethered fly, the steering trajectory is phase advanced relative to the moving bar, apparently in anticipation of its future position. By contrast, flickering bars that generate no coherent motion or have a surface texture that moves in the direction opposite to the bar generate steering responses that lag behind the stimulus. It remains unclear how the spatial properties of a bar influence behavioral response dynamics. Here, we show that a dark bar defined by its luminance contrast to the uniform background drives a co-directional steering response that is phase advanced relative to the response to a textured bar defined only by its motion relative to a stationary textured background. The textured bar drives an initial contra-directional turn and phase-locked tracking. The qualitatively distinct response dynamics could indicate parallel visual processing of a luminance versus motion-defined object. Calcium imaging shows that T4/T5 motion-detecting neurons are more responsive to a solid dark bar than a motion-defined bar. Genetically blocking T4/T5 neurons eliminates the phase-advanced co-directional response to the luminance-defined bar, leaving the orientation response largely intact. We conclude that T4/T5 neurons mediate a co-directional optomotor response to a luminance-defined bar, thereby driving phase-advanced wing kinematics, whereas separate unknown visual pathways elicit the contra-directional orientation response.

Local synaptic integration enables ON-OFF asymmetric and layer-specific visual information processing in vGluT3 amacrine cell dendrites

A basic scheme of neuronal organization in the mammalian retina is the segregation of ON and OFF pathways in the inner plexiform layer (IPL), where glutamate is released from ON and OFF bipolar cell terminals in separate inner (ON) and outer (OFF) sublayers in response to light intensity increments and decrements, respectively. However, recent studies have found that vGluT3-expressing glutamatergic amacrine cells (GACs) generate ON-OFF somatic responses and release glutamate onto both ON and OFF ganglion cell types, raising the possibility of crossover excitation in violation of the canonical ON-OFF segregation scheme. To test this possibility, we recorded light-evoked Ca2+ responses from dendrites of individual GACs infected with GCaMP6s in mouse. Under two-photon imaging, a single GAC generated rectified local dendritic responses, showing ON-dominant responses in ON sublayers and OFF-dominant responses in OFF sublayers. This unexpected ON-OFF segregation within a small-field amacrine cell arose from local synaptic processing, mediated predominantly by synaptic inhibition. Multiple forms of synaptic inhibition compartmentalized the GAC dendritic tree and endowed all dendritic varicosities with a small-center, strong-surround receptive field, which varied in receptive field size and degree of ON-OFF asymmetry with IPL depth. The results reveal a form of short-range dendritic autonomy that enables a small-field, dual-transmitter amacrine cell to process diverse dendritic functions in a stratification level- and postsynaptic target-specific manner, while preserving the fundamental ON-OFF segregation scheme for parallel visual processing and high spatial resolution for small object motion and uniformity detection.

The channel capacity of visual awareness divided among multiple moving objects

If attention is distributed among multiple moving objects, how does this divided attention affect the temporal process for detecting a specific target motion? Well-trained observers in three experiments monitored ongoing random motions of multiple objects, trying to rapidly detect non-random target motions. Response time hazard rates revealed a simple lawful structure of the detection processes. Target detection rates (hazard rates, in bits /s) were inversely proportional to the number of observed objects. Detection rates at any response time and in any condition equaled a product of two parallel (functionally independent and concurrent) visual processes: visual awareness and motion integration. The rate of visual awareness was inversely proportional to Set Size (n = 1-12), constant over time, and invariant with integrated motion information. Thus, a single rate parameter, indicating a constant channel capacity of visual awareness, described detection rates over a wide range of conditions and response times. During an initial interval of roughly 0.5s, detection rates increased proportionally with the duration and length of motion; but after this initial integration, detection rates were constant, independent of the time the target remained undetected. The relationship between the quantity of visual information and detection rates was simpler than anticipated by contemporary theories of attention, perception, and performance.

Towards a better understanding of parallel visual processing in human vision: Evidence for exhaustive analysis of visual information.

Most current models of visual processing propose that there are 2 main stages of visual processing, the first consisting of a parallel visual analysis of the scene and the second being a precise scrutiny of a few elements in the scene. Here, we present novel evidence that the first stage of processing adds systematic variance to visual processing times. When searching for a specific target, it has a behaviorally unique signature: RTs increase logarithmically with the number of items in the display and this increase is modulated by target-distractor similarity. This signature is characteristic of unlimited capacity parallel and exhaustive processing of all the elements in the scene. The function of this processing is to identify the locations in the scene containing items that are sufficiently similar to the target as to merit focused scrutiny, while discarding those that do not. We also demonstrate that stage-1 variability is sensitive to the observers' top-down goals: with identical displays, whereas RTs increase logarithmically with set size when observers are asked to find a specific target, they decrease exponentially when asked to find a unique item in the scene. (PsycINFO Database Record

Neural Mechanisms for Drosophila Contrast Vision

Spatial contrast, the difference in adjacent luminance values, provides information about objects, textures, and motion and supports diverse visual behaviors. Contrast computation is therefore an essential element of visual processing. The underlying mechanisms, however, are poorly understood. In human psychophysics, contrast illusions are means to explore such computations, but humans offer limited experimental access. Via behavioral experiments in Drosophila, we find that flies are also susceptible to contrast illusions. Using genetic silencing techniques, electrophysiology, and modeling, we systematically dissect the mechanisms and neuronal correlates underlying the behavior. Our results indicate that spatial contrast computation involves lateral inhibition within the same pathway that computes motion of luminance increments (ON pathway). Yet motion-blind flies, in which we silenced downstream motion-sensitive neurons needed for optomotor behavior, have fully intact contrast responses. In conclusion, spatial contrast and motion cues are first computed by overlapping neuronal circuits which subsequently feed into parallel visual processing streams.

SV3. Modulation of hemifield effects in visual-spatial attention by transcranial direct current stimulation

Introduction A bilateral field advantage (BFA), that is enhanced visual processing when stimuli are aligned across both visual hemifields, corresponds to a hemispheric resource model of parallel visual attentional processing, suggesting more attentional resources for bilateral displays. Several studies have shown that the BFA extends beyond early stages of visual attentional processing. Visual-short-term-memory (VSTM) capacity was higher when stimuli are distributed bilaterally rather than unilaterally and repetitive transcranial magnetic stimulation (rTMS) over the right precuneus diminished the BFA (Kraft et al., 2014). Methods: Here we tested whether the BFA can be modulated by simultaneous anodal and cathodal transcranial direct current stimulation applied either over the left and right occipital cortex (O1/O2) or over the left and right parietal cortex (P3/P4). Twenty healthy subjects were tested using a whole report paradigm based on the Theory of Visual Attention (TVA; Bundesen, 1990). Results and Conclusion: This approach allowed us measuring VSTM capacity and visual processing speed variability in unilateral and bilateral displays. It is discussed how this approach can modulate the inter-hemispheric parietal and occipital balance and whether it can be used in therapeutic settings, e.g. in patients with visual deficits or spatial neglect.

Abnormalities in brain systems supporting individuation and enumeration in autism.

Previous work indicates that adults with autism display a decreased capacity when rapidly enumerating small sets of elements (i.e., subitizing), compared to typically developing (TD) individuals. This ability is crucial for fundamental visual functions such as object individuation and parallel processing. Thus, the deficit in autism suggests limits in these skills. To examine the neural basis of this limitation, adults with and without high functioning autism rapidly enumerated 1 to 8 randomly located squares during a neuroimaging study. Typically, adults are thought to use parallel visual processes to quantify up to three or four elements, and serial processes to enumerate more (5+) elements. We hypothesized that parietal lobe regions associated with counting would be recruited with smaller sets of elements in adults with autism, compared to TD adults. Consistent with this hypothesis, activation in parietal regions increased with smaller set sizes in adults with autism compared to TD adults. Increased activation for three elements was evident in several regions, including those thought to underlie subitizing. In addition, regions specific to the counting range in TD adults were often equally active for set sizes in the subitizing range in the adults with autism. Finally, significant deactivation was evident in TD adults, presumably reflecting relative suppression of regions specialized for competing processes, but was not apparent in adults with autism. These differences in brain function in adults with autism on a simple enumeration task suggest atypical brain organization and function that is likely to impact most visual tasks, especially those with multiple elements.

TMS Over the Right Precuneus Reduces the Bilateral Field Advantage in Visual Short Term Memory Capacity

Several studies have demonstrated a bilateral field advantage (BFA) in early visual attentional processing, that is, enhanced visual processing when stimuli are spread across both visual hemifields. The results are reminiscent of a hemispheric resource model of parallel visual attentional processing, suggestingmore attentional resources on an early level of visual processing for bilateral displays [e.g. Sereno AB, Kosslyn SM. Discrimination within and between hemifields: a new constraint on theories ofattention. Neuropsychologia 1991;29(7):659-75.]. Several studies have shown that the BFA extends beyond early stages of visual attentional processing, demonstrating that visual short term memory (VSTM) capacity is higher when stimuli are distributed bilaterally rather than unilaterally. Here we examine whether hemisphere-specific resources are also evident on later stages of visual attentional processing. Based on the Theory of Visual Attention (TVA) [Bundesen C. A theory of visual attention. Psychol Rev 1990;97(4):523-47.] we used a whole report paradigm that allows investigating visual attention capacity variability in unilateral and bilateral displays during navigated repetitive transcranial magnetic stimulation (rTMS) of the precuneus region. A robust BFA in VSTM storage capacity was apparent after rTMS over the left precuneus and in the control condition without rTMS. In contrast, the BFA diminished with rTMS over the right precuneus. This finding indicates that the right precuneus plays a causal role in VSTM capacity, particularly in bilateral visual displays.

Distinct Contributions of the Magnocellular and Parvocellular Visual Streams to Perceptual Selection

During binocular rivalry, conflicting images presented to the two eyes compete for perceptual dominance, but the neural basis of this competition is disputed. In interocular switch rivalry, rival images periodically exchanged between the two eyes generate one of two types of perceptual alternation: (1) a fast, regular alternation between the images that is time-locked to the stimulus switches and has been proposed to arise from competition at lower levels of the visual processing hierarchy or (2) a slow, irregular alternation spanning multiple stimulus switches that has been associated with higher levels of the visual system. The existence of these two types of perceptual alternation has been influential in establishing the view that rivalry may be resolved at multiple hierarchical levels of the visual system. We varied the spatial, temporal, and luminance properties of interocular switch rivalry gratings and found, instead, an association between fast, regular perceptual alternations and processing by the magnocellular stream and between slow, irregular alternations and processing by the parvocellular stream. The magnocellular and parvocellular streams are two early visual pathways that are specialized for the processing of motion and form, respectively. These results provide a new framework for understanding the neural substrates of binocular rivalry that emphasizes the importance of parallel visual processing streams, and not only hierarchical organization, in the perceptual resolution of ambiguities in the visual environment.

Brake lamp detection in complex and dynamic environments: Recognizing limitations of visual attention and perception

Worldwide, both brake lamps and tail lamps on motor vehicles are required to be red. Previous studies have not examined the effect of this confound in a complex, high-traffic scenario in a driving simulator or on visuomotor behavior. In the first experiment, drivers detected brake lamps on nine lead vehicles and lane changes on two rear vehicles in a 15min simulated night time highway drive. A second experiment was used to examine the findings in the context of pre-attentive visual processing research. A third experiment analyzed visuomotor behavior and subjective workload during a vigilance task to further evaluate this hypothesis. For all studies, tail lamp color was manipulated, resulting in two conditions: the currently mandated red tail lamps and red brake lamps vs. yellow tail lamps and red brake lamps. Compared to current rear lighting, employing yellow tail lamps with red brake lamps reduced RT, error, subjective workload, improved performance in detecting lane changes and also changed visuomotor behavior. It is suggested that the mechanism allowing better performance is pre-attentive, parallel visual processing.

A visual processing but no phonological disorder in a child with mixed dyslexia

The case study of Martial, a French 9-year-old boy, who exhibits severe mixed dyslexia and surface dysgraphia is reported. Despite very poor pseudo-word reading, Martial has preserved phonological processing skills as his good oral language, good phoneme awareness and good verbal short-term memory show. He exhibited a strong length effect when reading briefly presented words but no sign of mini-neglect. His letter-string processing abilities were assessed through tasks of whole and partial report. In whole report, Martial could only name a few letters from briefly displayed 5-consonant strings. He showed an initial-position advantage and a sharper than expected left-to-right gradient of performance. He performed better when asked to report a single cued letter within the string but then showed an atypical right-side advantage. The same rightward attentional bias was observed in whole report when top-down control was prevented. Otherwise, Martial showed preserved single letter identification skills and good processing of 5-letter strings when letters were sequentially displayed one at a time. His poor letter-string processing thus reflects a parallel visual processing disorder that is compatible with either a visual attention (VA) span or a visual short-term memory disorder. Martial was further engaged in a complex reaching movement task involving VA and simultaneous processing. He performed motor sequences not as a whole but as a succession of independent motor units, suggesting that his attention was not allocated in parallel to the two to-be-reached targets prior to movement execution. Against a more basic motor disorder however, he showed good performance in a task of cyclical pointing movements. The overall findings suggest that Martial suffers from a visual simultaneous processing disorder that disturbs letter identification in strings. Instead of being restricted to letter-string processing, this VA disorder might extend to non-verbal task.

The top-down control of visual selection and how it is linked to the N2pc component

Is the spatial selection of visual objects fully under the control of top-down factors such as current tasks sets? In his admirably clear and systematic review (Theeuwes, 2010), Jan Theeuwes claims that the selection of visual objects is determined solely by bottom-up mechanisms that are independent of observers' selection intentions, and that top-down control only comes into play at later stages of attentional processing. His argument starts with the assumption that visual information is initially processed in a parallel and feedforward fashion, and that this early stage is entirely stimulus-driven. That much is not contentious — in fact, many models of visual processing and attentional selectivity assume the existence of an early fast feedforward sweep that is essentially non-selective. Theeuwes' central and controversial claim concerns the factors that drive the subsequent attentional selection of visual objects. He argues that this selection is determined by bottom-up salience maps that are computed during the initial parallel visual processing stage. Crucially, this salience-driven attentional object selection cannot be prevented or even modulated by top-down task set. Top-down factors can only influence what happens after an object has been selected; for example, attention can be rapidly disengaged from objects that do not possess currently task-relevant attributes. But intentional factors have no role whatsoever in the attentional selection of visual objects. It is unusual to define a psychological concept (bottom-up visual selection) by applying a negative criterion (the absence of intentional modulation). As a result, there is a surprisingly wide range of cases that appear to meet this criterion. For example, Theeuwes considers phenomena such as intertrial feature priming or memory-driven attentional capture to be instances of bottom-up selection, even though here the selection of visual objects is not determined by their salience. Such an extension of the concept of bottom-up selection

Action-Related Properties Shape Object Representations in the Ventral Stream

The principles driving the organization of the ventral object-processing stream remain unknown. Here, we show that stimulus-specific repetition suppression (RS) in one region of the ventral stream is biased according to motor-relevant properties of objects. Quantitative analysis confirmed that this result was not confounded with similarity in visual shape. A similar pattern of biases in RS according to motor-relevant properties of objects was observed in dorsal stream regions in the left hemisphere. These findings suggest that neural specificity for "tools" in the ventral stream is driven by similarity metrics computed over motor-relevant information represented in dorsal structures. Support for this view is provided by converging results from functional connectivity analyses of the fMRI data and a separate neuropsychological study. More generally, these data suggest that a basic organizing principle giving rise to "category specificity" in the ventral stream may involve similarity metrics computed over information represented elsewhere in the brain.