Complex Visual Environments Research Articles

Spatio-temporal target-measure association using an adaptive geometrical approach

Data association is of crucial importance to improve target tracking performance in many complex visual environments (non-linear dynamics, occlusions, etc). Usually, association effectiveness is based on prior information and observation category. However, association becomes difficult if targets are similar. Problems also arise in cases of missing data, complex motions or deformations over time. To remedy, we propose a new method for data association, that uses the evolution of the dynamic model of targets. The main idea is to measure an adaptive geometric accuracy between possible trajectories of targets, by only using positions as information, that constitutes its main advantage.

Real-time emulation of parallel channel responses in the vertebrate retina and the primary visual cortex

In order to examine the functional roles of visual neurons in the retina and the cortex, responses of neurons to visual stimuli have been investigated physiologically, and discussions have been made using models derived from the physiological experiments. However, these discussions do not necessarily applicable to the case of highly complex visual environments wherein actual biological vision systems operate since they are based on the responses of a limited number of neurons to simple visual stimuli. In order to examine and discuss the functional roles of the neurons, a system that can emulate the responses of a group of model neurons in real time is required. Thus far, a high-speed neural computation emulator with transient response has been developed mainly for engineering applications [1]. In this study, we have developed a system for the real-time reconstruction of neural activities with physiologically reasonable spatiotemporal properties using an analog-digital hybrid system comprised of analog resistive networks, field-programmable gate arrays (FPGA), and a digital computer. The system reconstructs activities of 128 x 128 neurons in each neural layer of the retina and the primary visual cortex at 200 frames per second. The system consists of two retina emulators, a cortex emulator, and a digital computer (Fig.1(A)). The retina emulator emulates both the sustained and transient responses (Fig.1(B)1(C)), and their spike representations (Fig.1(D)1(E)). A part of detailed operation of the retina emulator is written in [2]. The cortex emulator receives spike signals from two retina emulators and emulates responses of cortex neurons including the simple and complex cells. Figure 1 A. Appearance of the system comprised of two retina emulators, a cortex emulator, and a digital computer. B. Sustained response. C. Transient response. D. Sustained response with spike representation. E. Transient response with spike representation.

Real world referencing and schizophrenia: Are we experiencing the same reality?

Psychotic symptoms in schizophrenia patients encompass the difficulty to distinguish between the respective points of view of self and others. The capacity to adopt and switch between different perspectives is, however, fundamental for ego- and allocentric spatial referencing. We tested whether schizophrenia patients are able to adopt and maintain a non-egocentric point of view in a complex visual environment. Twenty-four chronic schizophrenic outpatients (11 females) and 25 controls matched for age, gender, years of education and handedness were recruited from a population-based sample. In a virtual environment, participants had to make a decision as to which of two trash cans was closest to themselves (viewer-centered, egocentric), to a ball (object-centered, unstable allocentric), or to a palace (landmark-centered, stable allocentric). Main outcome measures were reaction time, error rate, learning rate and local task switch cost. While egocentric reaction time was preserved, patients showed an increased reaction time in both allocentric referencing conditions (stable and unstable) and an overall increased error rate. Switch cost was diminished in patients when changing from the egocentric to the landmark-centered condition and elevated when changing from the landmark-centered to the egocentric condition. The results imply that schizophrenia patients’ adoption of an egocentric perspective is preserved. However, adopting an allocentric point of view and switching between egocentric and landmark-centered perspectives are impaired. Perturbations in non-egocentric referencing and transferring efficiently between different referential systems might contribute to altered personal and social world comprehension in schizophrenia.

Dynamic tracking re-adjustment: a method for automatic tracking recovery in complex visual environments

Detection and analysis of events from video sequences is probably one of the most important research issues in computer vision and pattern analysis society. Before, however, applying methods and tools for analyzing actions, behavior or events, we need to implement robust and reliable tracking algorithms able to automatically monitor the movements of many objects in the scene regardless of the complexity of the background, existence of occlusions and illumination changes. Despite the recent research efforts in the field of object tracking, the main limitation of most of the existing algorithms is that they are not enriched with automatic recovery strategies able to re-initialize tracking whenever its performance severely deteriorates. This is addressed in this paper by proposing an automatic tracking recovery tool which improves the performance of any tracking algorithm whenever the results are not acceptable. For the recovery, non-linear object modeling tools are used which probabilistically label image regions to object classes. The models are also time varying. The first property is implemented in our case using concepts from functional analysis which allow parametrization of any arbitrary non-linear function (with some restrictions on its continuity) as a finite series of known functional components but of unknown coefficients. The second property is addressed by proposing an innovative algorithm that optimally estimates the non-linear model at an upcoming time instance based on the current non-linear models that have been already approximated. The architecture is enhanced by a decision mechanism which permits verification of the time instances in which tracking recovery should take place. Experimental results on a set of different video sequences that present complex visual phenomena (full and partial occlusions, illumination variations, complex background, etc) are depicted to demonstrate the efficiency of the proposed scheme in proving tracking in very difficult visual content conditions. Additionally, criteria are proposed to objectively evaluate the tracking performance and compare it with other strategies.

The encoding process of nonconfigural information in contextual cuing

Jiang and Wagner (2004) demonstrated that individual target-distractor associations were learned in contextual cuing. We examined whether individual associations can be learned in efficient visual searches that do not involve attentional deployment to individual search items. In Experiment 1, individual associations were not learned during the efficient search tasks. However, in Experiment 2, where additional exposure duration of the search display was provided by presenting placeholders marking future locations of the search items, individual associations were successfully learned in the efficient search tasks and transferred to inefficient search. Moreover, Experiment 3 demonstrated that a concurrent task requiring attention does not affect the learning of the local visual context. These results clearly showed that attentional deployment is not necessary for learning individual locations and clarified how the human visual system extracts and preserves regularity in complex visual environments for efficient visual information processing.

Visual working memory for global, object, and part-based information

We investigated visual working memory for novel objects and parts of novel objects. After a delay period, participants showed strikingly more accurate performance recognizing a single whole object than the parts of that object. This bias to remember whole objects, rather than parts, persisted even when the division between parts was clearly defined and the parts were disconnected from each other so that, in order to remember the single whole object, the participants needed to mentally combine the parts. In addition, the bias was confirmed when the parts were divided by color. These experiments indicated that holistic perceptual-grouping biases are automatically used to organize storage in visual working memory. In addition, our results suggested that the bias was impervious to top-down consciously directed control, because when task demands were manipulated through instruction and catch trials, the participants still recognized whole objects more quickly and more accurately than their parts. This bias persisted even when the whole objects were novel and the parts were familiar. We propose that visual working memory representations depend primarily on the global configural properties of whole objects, rather than part-based representations, even when the parts themselves can be clearly perceived as individual objects. This global configural bias beneficially reduces memory load on a capacity-limited system operating in a complex visual environment, because fewer distinct items must be remembered.

Rapid discrimination of visual scene content in the human brain

The rapid evaluation of complex visual environments is critical for an organism's adaptation and survival. Previous studies have shown that emotionally significant visual scenes, both pleasant and unpleasant, elicit a larger late positive wave in the event-related brain potential (ERP) than emotionally neutral pictures. The purpose of the present study was to examine whether neuroelectric responses elicited by complex pictures discriminate between specific, biologically relevant contents of the visual scene and to determine how early in the picture processing this discrimination occurs. Subjects ( n = 264) viewed 55 color slides differing in both scene content and emotional significance. No categorical judgments or responses were required. Consistent with previous studies, we found that emotionally arousing pictures, regardless of their content, produce a larger late positive wave than neutral pictures. However, when pictures were further categorized by content, anterior ERP components in a time window between 200 and 600 ms following stimulus onset showed a high selectivity for pictures with erotic content compared to other pictures regardless of their emotional valence (pleasant, neutral, and unpleasant) or emotional arousal. The divergence of ERPs elicited by erotic and non-erotic contents started at 185 ms post-stimulus in the fronto-central midline region, with a later onset in parietal regions. This rapid, selective, and content-specific processing of erotic materials and its dissociation from other pictures (including emotionally positive pictures) suggests the existence of a specialized neural network for prioritized processing of a distinct category of biologically relevant stimuli with high adaptive and evolutionary significance.

An elaborated model of fly small-target tracking

Flies have the capability to visually track small moving targets, even across cluttered backgrounds. Previous computational models, based on figure detection (FD) cells identified in the fly, have suggested how this may be accomplished at a neuronal level based on information about relative motion between the target and the background. We experimented with the use of this "small-field system model" for the tracking of small moving targets by a simulated fly in a cluttered environment and discovered some functional limitations. As a result of these experiments, we propose elaborations of the original small-field system model to support stronger effects of background motion on small-field responses, proper accounting for more complex optical flow fields, and more direct guidance toward the target. We show that the elaborated model achieves much better tracking performance than the original model in complex visual environments and discuss the biological implications of our elaborations. The elaborated model may help to explain recent electrophysiological data on FD cells that seem to contradict the original model.

Target selection in area V4 during a multidimensional visual search task.

Natural scenes typically contain multiple objects that are unique in different stimulus dimensions so that an object with feature contrast to surrounding objects draws attention and pops out. Furthermore, if we have previous knowledge about the dimension in which a target object differs from the surrounding objects, we will attend to that dimension and more easily detect the target. Our aims here were to elucidate neural mechanisms underlying this type of attention by recording neuronal activities from area V4 and to investigate how visual signals encoding feature contrast between objects are modulated by attention specific to a particular dimension. To accomplish this, we trained monkeys to do a multidimensional visual search task in which two singleton stimuli, unique in the color or shape dimension, were presented with four other identical stimuli. The monkeys had to search for the singleton stimulus that was unique in the instructed dimension while the search dimension was switched between shape and color. We found that individual V4 neurons carry visual signals encoding feature contrast in either shape or color, and this signal is modulated depending on the search dimension. Population responses to the target singleton stimulus were significantly higher than to others, regardless of the search dimension. In most V4 neurons, however, significant response increases occurred only when one particular singleton stimulus was the target. These findings suggest that interaction between bottom-up signals encoding feature contrast between stimuli and top-down signals encoding search dimension occurs in V4 and facilitates adaptive selection of targets in a complex visual environment.

Driver distraction by advertising: genuine risk or urban myth?

Drivers operate in an increasingly complex visual environment, and yet there has been little recent research on the effects this might have on driving ability and accident rates. This paper is based on research carried out for the Scottish Executive's Central Research Unit on the subject of external-to-vehicle driver distraction. A literature review/meta-analysis was carried out with a view to answering the following questions: is there a serious risk to safe driving caused by features in the external environment, and if there is, what can be done about it? Review of the existing literature suggests that, although the subject is under-researched, there is evidence that in some cases overcomplex visual fields can distract drivers and that it is unlikely that existing guidelines and legislation adequately regulate this. Theoretical explanations for the phenomenon are offered and areas for future research highlighted.

Driver distraction by advertising: genuine risk or urban myth?

Drivers operate in an increasingly complex visual environment, and yet there has been little recent research on the effects this might have on driving ability and accident rates. This paper is based on research carried out for the Scottish Executive's Central Research Unit on the subject of external-to-vehicle driver distraction. A literature review/meta-analysis was carried out with a view to answering the following questions: is there a serious risk to safe driving caused by features in the external environment, and if there is, what can be done about it? Review of the existing literature suggests that, although the subject is under-researched, there is evidence that in some cases overcomplex visual fields can distract drivers and that it is unlikely that existing guidelines and legislation adequately regulate this. Theoretical explanations for the phenomenon are offered and areas for future research highlighted.

Color and defective color vision as factors in the conspicuity of signs and signals.

The conspicuity of road traffic signs and signals for a group of observers with the color vision defect of deuteranopia is compared with that for a control group of observers with normal color vision. Conspicuity was measured by the proportion of reports of target objects detected in 300-ms presentations of projected slides of road scenes. There were two instructions, one designed to measure attention conspicuity and the other, search conspicuity. The attention conspicuity of red, orange, and green color-coded traffic control devices was significantly less for deuteranopes than for the observers with normal color vision, but this was not true for yellow and blue color-coded signs. This result is consistent with our understanding of the color perceptions of deuteranopes. The reduction of conspicuity was not so great for the search conspicuity condition. We conclude that redundant color coding does contribute to the conspicuity of signs and signals and that deuteranopes--and probably those with other severe forms of defective color vision--have a significantly reduced ability to notice colored targets, such as road signs and signals, in complex visual environments. The actual and potential application of this work is in the design of signs so they are conspicuous, especially when the user group includes people with defective color vision.

Functional Anatomy of Visual Search: Regional Segregations within the Frontal Eye Fields and Effective Connectivity of the Superior Colliculus

The ability to find targets embedded within complex visual environments requires the dynamic programming of visuomotor search behaviors. Functional magnetic resonance imaging was used to image subjects while they visually searched for targets embedded among foils. Visuomotor search activated the posterior parietal cortex and the frontal eye fields. Both regions showed a greater number of activated voxels on the right, consistent with the known pattern of right hemispheric dominance for spatial attention. The superior colliculus showed prominent activation in the search versus eye movement contrast, demonstrating, for the first time in humans, activation of this region specifically related to an exploratory attentional contingency. An analysis of effective connectivity demonstrated that the search-dependent variance in the activity of the superior colliculus was significantly influenced by the activity in a network of cortical regions including the right frontal eye fields and bilateral parietal and occipital cortices. These experiments also revealed the presence of a mosaic of activated sites within the frontal eye field region wherein saccadic eye movements, covert shifts of attention, and visuomotor search elicited overlapping but not identical zones of activation. In contrast to the existing literature on functional imaging, which has focused on covert shifts of spatial attention, this study helps to characterize the functional anatomy of overt spatial exploration.

Using a mobile robot to study locust collision avoidance responses.

The visual systems of insects perform complex processing using remarkably compact neural circuits, yet these circuits are often studied using simplified stimuli which fail to reveal their behaviour in more complex visual environments. We address this issue by testing models of these circuits in real-world visual environments using a mobile robot. In this paper we focus on the lobula giant movement detector (LGMD) system of the locust which responds selectively to objects which approach the animal on a collision course and is thought to trigger escape behaviours. We show that a neural network model of the LGMD system shares the preference for approaching objects and detects obstacles over a range of speeds. Our results highlight aspects of the basic response properties of the biological system which have important implications for the behavioural role of the LGMD.

Coordination of arm movements in a complex visual environment

Coordination of arm movements in a complex visual environment

Metabolic activity in the hyperstriatum of 2-day-old chicks during optomotor and contrasting visual stimulation

Our earlier report of differences in metabolic activity within the visual regions of the hyperstriatum and ectostriatum, in 2-day-old chicks compared with 23-day-old chicks, suggested that two visual pathways within the visual system develop at different rates. Here we have investigated whether the demands of varying visual environments will increase the activity of the hyperstriatum accessorium (HA) in 2-day-olds. Metabolic activity in the HA was monitored in 2-day-old chicks by the radioactive 2-deoxyglucose technique during monocular stimulation with three different visual environments: moving stripes in a rotating drum, which induced eye and head movements, a featureless white environment, and the complex visual environment of the home cage with other chicks. Although a small but significant level of activity was found in HA in the hemisphere opposite the open eye, the activity did not vary with the visual treatment. On the other hand, a raised level of activity in the hyperstriatum dorsale (HD) appeared in chicks viewing the rotating stripes, indicating that at this age the thalamo-hyperstriatal pathway may be involved in processing whole-field visual movement. The optomoter environment also produced high activity in the medial hyperstriatum ventrale (MHV), a region that has been implicated in memory formation of imprinting. We suggest that during the sensitive period for imprinting, HA may either have not developed its fully functional capacity, or that following or during imprinting it is actively shut down to protect itself and associated regions from interfering visual input. In contrast to the 2-day-olds, 17-day-old chicks in a visually rich cage environment, had high levels of activity in HA, demonstrating that the functional maturation of the HA, related to performance in the cage environment, is complete at least 6 days earlier than previously observed.

A framework for object recognition in a visually complex environment and its application to locating address blocks on mail pieces

A computational framework for recognizing an object of interest in a complex visual environment is developed. Arising from the problem of finding the destination address block on a mail piece, a general framework for coordinating a collection of specialized image analysis tools is described. The framework is based on the blackboard model of problem-solving with enhancements to account for object recognition in a visually complex environment. Those enhancements include (i) a systematic derivation of knowledge sources that describes a methodology for developing knowledge sources in an application domain, (ii) a dependency graph to specify the interdependency between knowledge sources, (iii) a control mechanism that detects the varying degrees of complexity in the input, and adjust the behavior of the system so that the effort expended in gathering evidence is in accordance with the complexity of inputs, (iv) a formalism for estimating the utility of a tool so that the effective knowledge sources get high priorities to be invoked, (v) the organization of a knowledge source which allows the image analysis related knowledge sources to be reapplied on unsatisfactory initial results, and (vi) an evidence combination scheme to combine confidence values associated with pieces of evidence generated by various knowledge sources. An address block location system (ABLS) is developed and implemented in accordance with this framework and shown to be capable of dealing with a wide range of environments: from those having a high degree of global spatial structure (e.g., letter mail envelopes which conform to specifications) to those with no structure (e.g., magazines with randomly pasted address labels). Experimental results with a database of difficult cases demonstrating the promise of the computational framework are presented.

Traffic signs in complex visual environments

The effects of sign luminance on detection and recognition of traffic control devices is mediated through contrast with the immediate surround. Additionally, complex visual scenes are known to degrade visual performance with targets well above visual threshold. A laboratory study was conducted to determine ways of measuring visual complexity and to assess the capability of changes in sign luminance to offset decrements in performance resulting from added complexity. Positive results were found for yellow warning, construction, and stop signs but not for the black on white regulatory sign. Regression equations using complexity factors, contrast, and target variables suggested that in complex scenes, complexity is a more significant determinant of sign detection than brightness or contrast. A field study also was conducted to determine if these findings could be observed in real world driver performance. The effects of visual complexity were observed in the field and increasing sign brightness improved sign recognition under specific conditions. (FHWA)

Discrimination between movements of eye and object by visual interneurones of crickets.

ABSTRACT One large neurone on each side of the cervical and thoracic ventral nerve cord of crickets responds to object motion anywhere in the visual field of the ipsilateral compound eye, but not to the forced or voluntary movement of the eye itself. This discrimination between self-movement and object-movement is accomplished by an inhibitory mechanism mediated by the same eye. Inhibition must be present because a potent moving stimulus becomes ineffective if presented during a forced eye movement. Its visual origin is demonstrated in two ways : (a) abolishing all known mechano-sensory feedback does not disrupt the mechanism, but (b) alteration of visual conditions does so in a predictable way. Sweeping the eye past a complex visual environment suppresses the neurone’s response to a concurrently or subsequently presented moving target, whereas the same movement past a simplified or homogeneous environment produces little or no inhibition. Responses to eye movement itself are greatly enhanced in appropriately simplified visual fields, reinforcing the conclusion that the inhibition preventing response in complex fields is of visual origin. Suggestive evidence for an additional inhibitory mechanism associated with voluntary movement is presented.