Downloads & Free Reading Options - Results

A series of experiments are described which were conducted to further refine experimental paradigms for the investigation of information processing skills relevant to pilot training. A series of tasks have been developed and studied which attempt to measure the individual's information processing capacity as well as his susceptibility to performance degradation resulting from the introduction of interfering stimuli. Data suggest performance on these tasks to be highly dependent on individual differences, therefore, making them good candidates for use as tools in the investigation of information processing skills in flying training. Implications for direct application to flying training research are discussed. (Author/DGC)

One of a series of investigations on the Project on an Information Memory Model, the purpose of this study was to determine the amount and kind of visual information processed and stored in the memory of children using different modalities of observation. Children, aged 5, 9 and 13 years, were randomly assigned to one of three treatment groups. Two of the treatment groups individually viewed silently (N=21 per age level) or vocally (N=21 per age level) a projected 35mm slide of a living room. The third group (N-45 per age level) viewed the picture display in groups of three and verbally exchanged enumerations of items. Five minutes were allowed for viewing the slide. The subjects then verbally recalled the viewing experience and their recollections were tape recorded. As the age of the subjects increased, the numbe of picture item messages and variety or number of different picture items verbalized in the enumeration phase of the experiment also increased. A message processor algorithm for forecasting cognitions of verbal and nonverbal modalities was used with a prediction error of less than two percent. An item dimension algorithm for forecasting cognitions was developed to measure the M unit structure of memory pathways. The finding was that chronological age was related to M units of variations of learning and recall tasks. (Author/PEB)

This study aims to explore wherther and how tasting actual drinks would affect visual information processing.

This study aims to explore wherther and how tasting actual drinks would affect visual information processing.

The visual detection of a concealed target involves the interaction of object characteristics and observer-related variables. This report briefly summarises object characteristics which may affect detectability, and then concentrates upon observer variables. Some of the physical limitations of the human optical system are mentioned. An information processing theory is reviewed and adopted as a model upon which some hypotheses are based. In the case of a search of a large, complex field, automatic processing of information may lead to the selection of likely target areas, however controlled search processes may be required to scan the information in an automatically selected array.

Reported were four experiments which investigated the developmental and mental retardation aspects of an initial stage of visual information processing termed iconic memory. The stage was explained to involve processing visual stimuli prior to sensation and through to recognition. In three of the four experiments, the paradigm of visual masking was used to determine the speed or efficiency of information processing; the Crawford-pattern mask technique was used throughout. The four experiments were titled age effects on backward visual masking, adult age effects on backward visual masking, forward and backward masking with normal and retarded subjects, and iconic storage and information availability by normal and retarded subjects, respectively. Subjects in the four experiments ranged in age from 5 to 55 years of age. Major findings of the experiments were reported to be development and documentation of a methodology surmounting previous problems in studying development and mental retardation aspects of iconic memory, establishment of relationship between age variable and efficiency with which visual information is processed, and poorer performance by retarded subjects on iconic memory tasks than by normal subjects matched in terms of mental age. (CB)

This final progress report summaries the main recent results; full reports of the results are contained in the papers appended herewith. The summary also reviews some results from previous AFOSR grants where these are necessary to provide the background for the current research. Four areas are summarized: (1) Basic Mechanisms of Visual Motion and Texture Perception; (2) Lateral Interations in Texture Stimuli; (3) Information Processing; and (4) Visual Attention and Short-Term Memory.

The simultaneous performance of two single-dimensional compensatory tracking tasks, one with the left hand and one with the right hand, is discussed. The tracking performed with the left hand was considered the primary task and was performed with a visual display or a quickened kinesthetic-tactual (KT) display. The right-handed tracking was considered the secondary task and was carried out only with a visual display. Although the two primary task displays had afforded equivalent performance in a critical tracking task performed alone, in the dual-task situation the quickened KT primary display resulted in superior secondary visual task performance. Comparisons of various combinations of primary and secondary visual displays in integrated or separated formats indicate that the superiority of the quickened KT display is not simply due to the elimination of visual scanning. Additional testing indicated that quickening per se also is not the immediate cause of the observed KT superiority.

Researchers are concerned with the end-to-end performance of image gathering, coding, and processing. The applications range from high-resolution television to vision-based robotics, wherever the resolution, efficiency and robustness of visual information acquisition and processing are critical. For the presentation at this workshop, it is convenient to divide research activities into the following two overlapping areas: The first is the development of focal-plane processing techniques and technology to effectively combine image gathering with coding, with an emphasis on low-level vision processing akin to the retinal processing in human vision. The approach includes the familiar Laplacian pyramid, the new intensity-dependent spatial summation, and parallel sensing/processing networks. Three-dimensional image gathering is attained by combining laser ranging with sensor-array imaging. The second is the rigorous extension of information theory and optimal filtering to visual information acquisition and processing. The goal is to provide a comprehensive methodology for quantitatively assessing the end-to-end performance of image gathering, coding, and processing.

Eye movements place a limit on the processing of visual information because they determine the location and the velocity of the retinal image. Thus, to understand how we see it is necessary to understand how eye moments are controlled. Work this year in my laboratory has concentrated on the roles of expectations and selective attention in the programming of smooth and saccadic eye movements. We have: 1) demonstrated distinct roles for past experience and expectations in the control of smooth eye movement and found that expectation will predominate in the presence of cues about the direction of future motion; 2) found that brisk initial pursuit requires that expectation that target motion will continue; 3) showed that saccadic eye movements are not attracted to visual backgrounds (as had been claimed) unless subjects pay attention to the background. These studies show that central representations of visual scenes, containing information about the position, motion and future motion of selected objects, are the natural effective stimulus for human eye movement. Psychomotor tests; Psychology; Communication; Eye movement.

When we look around a natural environment the eye is directed toward objects we select. We seem to do this task accurately and effortlessly. Yet, even such a simple task presents real problems for the oculomotor system, namely, to select the relevant target from the detailed background so that only information contained in it influences the line of sight and to spatially-pool information in the selected target so that the line of sight lands at a single position within the selected target. We have found that: (1) the saccadic target is designated by means of selective perceptual attention, which means people cannot prepare to look to one target while accurately perceiving targets elsewhere, and (2) there is a highly-accurate spatial pooling process which can direct the line of sight to precise positions within large targets. The results show that the oculomotor system is capable of extremely rapid and effective scanning. The procedures humans use to accomplish this task may prove useful for the guidance of robotic systems which need to move about in patterned visual environments.

This report used a combination of physiological and anatomical approaches to elucidate the functional organization of visual cortex in the macaque monkey. One project was a single cell analysis of texture vision, using texture patterns of the type developed by Julesz for human psychophysical studies. Many cells tested in area V2 responded to static or moving texture gradients in ways which were not predictable on the basis of responses to individual texture elements and which correlated with the preattentive discriminability of these texture patterns to human observers. A second project involved the development of a computerized technique for generating two-dimensional maps of cerebral cortex. An algorithm based on simulated annealing procedures was used to construct a complete map of primary visual cortex, thereby demonstrating its suitability for dealing with anatomical data from highly convoluted regions of cortex. A third project involved the use of voltage-sensitive dyes to monitor activity patterns in visual cortex. This technique offers great promise for analyzing the organization of large neural ensembles with high spatial and temporal resolution. Keywords: Cerebral cortex; Pattern recognition; Cortical mapping; Single neuron physiology.

Limitations in the ability to temporarily represent information in visual working memory (VWM) are crucial for (visual) cognition. Whether VWM processing is dependent on an object’s saliency (i.e., how much it stands out) has been neglected in VWM research. Therefore, we developed a novel VWM task design that allows direct control over saliency. In three experiments with this task (on 10, 31, and 60 healthy adults, respectively), we consistently find that VWM performance is strongly and parametrically influenced by saliency and that both an object’s relative (to concurrently presented objects) and absolute saliency influence VWM processing. We also demonstrate that this effect is indeed due to bottom-up saliency, rather than differential fit between each object and the top-down attentional set. A simple computational model assuming that VWM performance is determined by the weighted sum of absolute and relative saliency accounts well for the observed data patterns.

This project examined the effects menstruation may have on visual attention in women. A recent study examined electroencephalographic (EEG) gender differences using a visual object recognition task. Results indicated certain EEG amplitudes (specifically, P300 and N400) are greater in women than men. This study extended the previous findings to determine if these increased EEG amplitudes vary across menstrual phases. Eighteen female participants participated in a series of 3 EEG recording sessions using the same visual object recognition task from the previous study; 18 male participants completed this task once. Analyses from 15 of the 18 female and 16 of the 18 male participants support the previous finding of larger P300 amplitudes in response to relevant stimuli for women compared with men. While there was no distinctive N400 component in this study, there was a late negative (LN) component which was found to vary significantly between men and women. In addition, multiple visual evoked potential (VEP) components varied significantly across the menstrual cycle. In particular, the N200 component appeared to provide greater differences between menstrual phases than either the p300 or LN components; however, the results varied greatly by head location, The differentiation found with VEP components in response to the pop-out task used in this study provide support for basic visual processing variation across the menstrual cycle and between genders.

Includes bibliographies

Functional organization of visual fields

Functional organization of visual fields

Memory and Cognition : v.9, no.3, 1981, p. 301-307

The central goal of the project funded by the current grant is the understanding of how saccadic eye movements are used to accurately localize objects in space. Explaining accurate localization is a real problem because the objects we choose to look at are relatively large -- typically extending over several degrees of space -- but the saccadic eye movement must land at one place within the target. Our published research suggests that the computation requires a sequence of two stages: (1) voluntary attentional weighting of information in the visual display, and (2) automatic-spatial-pooling of the weighted formation (He and Kowler, 1989; 1991).

Completed studies showed that: 1) saccade-like stimulus transients (abrupt onsets or off-sets) do not aid visual search; 2) anticipatory smooth eye movements depend in a lawful way on the stimuli in prior trials; 3) smooth pursuit eye movements are determined by two, independent processes; 4) reading efficiency is not limited by the pattern of eye movements, but rather by the ability to recognize words quickly; 5) subjects can maintain the line of sight on one of two, superimposed, full-field, patterns of randomly positioned dots -- one pattern moving and the other stationary.

A continuing problem in naval aviation is how to identify, prior to training, those individuals having the greatest likelihood to succeed in flight training. Since World War II the US Navy has been concerned with the development of selection tests to improve its capability to screen individuals desiring entry into flight training. Improved aviation personnel selection tests could reduce attrition, and consequently, the monetary losses associated with this attrition. Additionally, the Navy would expose fewer flight students to a potentially hazardous training situation. A research methodological problem associated with the development of candidate tests for predictive validation is: How do we design a test that measures abilities relevant for inclusion in validation studies? A methodology, Random Sampling of Domain Variances, was developed as a product of the Augmentation of Human Factors Engineering Technology Efforts Program. This methodology provided the basis for developing a complex visual information task designed to require many basic cognitive processing demands thought to be required in performing aviation display tasks. This complex visual display task should be further evaluated in a cross- validation study and then considered for implementation as a performance-based selection test. The method of Random Sampling of Domain Variances should be considered as a methodological tool in the design of future tests to be selected for validation studies. (sdw)

Tri 1 2019/2020

This study uses the 7.5% carbon dioxide model of anxiety induction to investigate the effects of state anxiety on visual and auditory information processing. Participants will undergo two inhalations (of 7.5% CO2-enriched air and medical air) and complete two tasks that present clear and degraded visual and auditory stimuli. We hypothesise that 7.5% CO2 inhalation will be associated with impaired performance and this effect will be greatest when stimuli are degraded.

Two experiments investigated a change in the late positive activity (200 to 550 msec) of the visual evoked potential (VEP). This 'P2-effect,' maximal at the occiput, appeared related to cognitive aspects of information processing. Results did not support an interpretation of the P2-effect in cognitive terms (Experiment 1) but rather in terms of shifts in cortical arousal and in expectancy (Experiment 2). In both studies, subjects viewed letter-pair stimuli, e.g., HH or bD, responding when they saw a stimulus (SIGHT task), or when they determined whether the two letters were the same or different in either size or name (SIZE and NAME tasks). Reaction times increased from the SIGHT to the SIZE to the NAME task, objectifying the increase in task complexity. The P2-effect occurred with the increase in task complexity from the SIGHT to the SIZE task, but did not increase further in the NAME task or show asymmetry in relation to lateralized processing. In Experiment 2, slow wave shifts were observed. These slow waves revealed subject-specific ranges of cortical arousal. The P2-effect appeared to result primarily from continued or noncontinued expectancy following stimulus reception. This work demonstrates the difficulty involved in controlling for and evaluating the effects of shifts in cortical arousal and in expectancy during cognitive information processing. The results are relevant to the question of CNV-P300 dissociability.

Food information has a high reward value for human beings, and may be more so for those who are overweight or obese. Stimuli with high reward value may be particularly conspicuous, demanding our attention even when we are attending to something else. Yet, the nature and neural signals of potential biased processing of food information remains unclear. Moreover, despite decades of research, a reliable measure of food information processing remains elusive, making it a methodological bottleneck in obesity and disordered eating research. To address the nature of biased attention processing, we ask 1) how attention is biased toward two cardinal properties of food information, namely, healthfulness and tastiness, even though the food information is irrelevant to the task at hand; and 2) whether and how such processing might be dependent on the demandingness of the central task (i.e., perceptual load). In our previous study which has been preregistered, we recruited 32 participants and measured their responses to infrequent visual food information embedded in streams of other visual objects in EEG, specifically, by using a variant of steady-state visual evoked potential (SSVEP), oddball fast periodic visual stimulation (FPVS). The infrequent food objects (i.e., oddballs) are one of four different combinations in different trials based on healthfulness and tastiness: 1) high healthfulness and high tastiness; 1) high healthfulness and low tastiness; 3) low healthfulness and high tastiness; and 4) low healthfulness and low tastiness. The perceptual load of the central task is either low (detection of a certain color of letters) or high (detection of certain combinations of color and identity of letters). In this study, the results demonstrated robust and reliable implicit categorization of food information at the individual level for all 32 participants, with the effect not modulated by perceptual load or tastiness (but might be by healthfulness). Since oddball FPVS is a novel paradigm applied to studying food information processing, it would be critical to replicate the findings in order to develop FPVS as a reliable measure of food information processing as well as apply this tool to obesity or other clinical sample with eating disorders. In particular, the modulating effect of healthfulness observed was not salient and requires further evidence. Furthermore, the lack of modulation by tastiness and perceptual load also warrants replication. Therefore, we plan to conduct a replication study with the same paradigm and sample size to validate our previous findings.

This research extended our understanding of the visual and cognitive process controlling saccadic and smooth eye movements, and the role of these eye movements in visual information acquisition. Experiments showed that: (1) saccades are biased toward likely locations of targets, suggesting that previous reports of 'center-of-gravity' reflexes are actually due to search of attentional strategies; (2) saccades can be directed to spatially-extended targets with an accuracy and precision as good as those found for single point targets; (3) predictive smooth eye movements are caused by cognitive expectations about future path of target motion, not by learned oculomotor habits; (4) slow control is not sensitive to position error; (5) smooth eye movements are sensitive to the expected direction of future target motion; (6) strategies of scanning the boundaries of difficult texture patterns are more effective than strategies of scanning the symmetric axis; (7) normal reading is carried out by a coordinated pattern of eye movements and head movements.

Chapter 1 of this report, "Introduction and General Recommendations for Eye-Movement Research and Instrumentation," discusses research priorities; encouraging and supporting theories, models, or simulations of information processing; and improved instrumentation in the field of visual information processing. Chapter 2, "Summary of Specific Recommendations," presents specific recommendations for the following four clusters: (1) reading texts, pictures, and other graphic displays; (2) processing television displays; (3) eye-movement research; and (4) instrumentation and instrumentalities. Chapter 3, "Specific Recommendations for Directions in Eye-Movement Research," recommends eye-movement research that explores the moment-to-moment information intake and processing mechanisms in the encounter of an individual with text, pictures, and other static graphic display; suggests research that explores the intake and processing of dynamic visual information as it is received from television; and argues the general value of eye movement research. Chapter 4, "Instrumentation and Instrumentalities," has four parts relating to the following areas: research, management, review, and support; instrumentation development; facilities dissemination and establishment; and priorities. (WR)

Information transfer between the operator and computer-generated display systems is an area where the human factors engineer discovers little useful design data relating human performance to system effectiveness. This study utilized a computer-driven, cathode-ray-tube graphic display to quantify human response speed in a sequential information processing task. The performance criteria was response time to sixteen cell elements of a square matrix display. A stimulus signal instruction specified selected cell locations by both row and column identification. An equal probable number code, from one to four, was assigned at random to the sixteen cells of the matrix and correspondingly required one of four, matched keyed-response alternatives. The display format corresponded to a sequence of diagnostic system maintenance events, that enable the operator to verify prime system status, engage backup redundancy for failed subsystem components, and exercise alternate decision-making judgements. The experimental task bypassed the skilled decision-making element and computer processing time, in order to determine a lower bound on the basic response speed for given stimulus/response hardware arrangement.

Information transfer between the operator and computer-generated display systems is an area where the human factors engineer discovers little useful design data relating human performance to system effectiveness. This study utilized a computer-driven, cathode-ray-tube graphic display to quantify human response speed in a sequential information processing task. The performance criteria was response time to sixteen cell elements of a square matrix display. A stimulus signal instruction specified selected cell locations by both row and column identification. An equal probable number code, from one to four, was assigned at random to the sixteen cells of the matrix and correspondingly required one of four, matched keyed-response alternatives. The display format corresponded to a sequence of diagnostic system maintenance events, that enable the operator to verify prime system status, engage backup redundancy for failed subsystem components, and exercise alternate decision-making judgements. The experimental task bypassed the skilled decision-making element and computer processing time, in order to determine a lower bound on the basic response speed for given stimulus/response hardware arrangement.

Differential levels of automaticity for processing information from tachistoscopic displays were studied using three variations of the partial report procedure. Processing of both spatial and identity information in letter displays was assessed simultaneously and then independently. Analyses indicated feature extraction is a highly automated process while spatial information extraction is a controlled process. Implications for separate processing pathways in the visual system are discussed

Visual processing of televised information was compared among 85 Japanese and 111 American boys and girls at the kindergarten and 4th-grade levels. The literatures on cognition and learning indicate that language and child rearing factors are more conducive to the development of iconic processing skills in Japanese children than in American children, a phenomenon which increases with age. The reverse pattern is found for verbal processing skills. It was predicted that after viewing a television program without narration or dialogue, Japanese children's amount and concentration of visual attention and comprehension would be higher than that of American children, particularly at the 4th-grade level. Results confirmed the predictions for visual attention, suggesting that different television processing strategies were employed. Contrary to the predictions, American children scored higher than Japanese children on the comprehension test, perhaps because of the cultural differences in expectations about the testing procedure. The results are discussed in the context of processing differences. Alternative explanations for the findings are explored, and directions for future research are suggested. Appended are seven pages of references, seven tables and two figures. (Author/RH)

This project concerned the discovery and description of basic mechanisms of human visual motion and texture perception. Motion and texture are critical inputs to visual perception. Basic mechanisms of motion are of particular interest because they are perhaps the primary substrate for perceptual recovery of 3D depth structures and orientation in space, they are critical for detecting new objects and events in the environment, as well as playing an important role in 2D perception. Motion and texture are considered together here because the problem of discriminating velocity in a one- dimensional motion stimulus is formally equivalent to the problem of discriminating orientation in a texture stimulus: the t dimension of the motion stimulus becomes the y dimension of the texture stimulus.

This project concerned the discovery and description of basic mechanisms of human visual motion and texture perception. Motion and texture are critical inputs to visual perception. Basic mechanisms of motion are of particular interest because they are perhaps the primary substrate for perceptual recovery of 3D depth structures and orientation in space, they are critical for detecting new objects and events in the environment, as well as playing an important role in 2D perception. Motion and texture are considered together here because the problem of discriminating velocity in a one- dimensional motion stimulus is formally equivalent to the problem of discriminating orientation in a texture stimulus: the t dimension of the motion stimulus becomes the y dimension of the texture stimulus.

This article is from Frontiers in Human Neuroscience , volume 7 . Abstract Objective: Mismatch negativity (MMN) and P3 are unique ERP components that provide objective indices of human cognitive functions such as short-term memory and prediction. Bipolar disorder (BD) is an endogenous psychiatric disorder characterized by extreme shifts in mood, energy, and ability to function socially. BD patients usually show cognitive dysfunction, and the goal of this study was to access their altered visual information processing via visual MMN (vMMN) and P3 using windmill pattern stimuli.Methods: Twenty patients with BD and 20 healthy controls matched for age, gender, and handedness participated in this study. Subjects were seated in front of a monitor and listened to a story via earphones. Two types of windmill patterns (standard and deviant) and white circle (target) stimuli were randomly presented on the monitor. All stimuli were presented in random order at 200-ms durations with an 800-ms inter-stimulus interval. Stimuli were presented at 80% (standard), 10% (deviant), and 10% (target) probabilities. The participants were instructed to attend to the story and press a button as soon as possible when the target stimuli were presented. Event-related potentials (ERPs) were recorded throughout the experiment using 128-channel EEG equipment. vMMN was obtained by subtracting standard from deviant stimuli responses, and P3 was evoked from the target stimulus.Results: Mean reaction times for target stimuli in the BD group were significantly higher than those in the control group. Additionally, mean vMMN-amplitudes and peak P3-amplitudes were significantly lower in the BD group than in controls.Conclusions: Abnormal vMMN and P3 in patients indicate a deficit of visual information processing in BD, which is consistent with their increased reaction time to visual target stimuli.Significance: Both bottom-up and top-down visual information processing are likely altered in BD.

The research completed over the last twelve months has included a number of studies concerned with evoked cortical potential correlates of visual stimulus processing in humans. The first experiment was conducted in order to replicate an earlier finding in which the amplitude of a relatively late positive component of the event related potential (ERP), known as the P3, was larger to a line orientation that required a 'yes' response than one that required a 'no'. The angular difference between the two lines was only 5 degrees. Another purpose was to assess the reliability of the better right hemisphere performance found with male subjects in the line orientation discrimination task used. In a second experiment, we examined the possibility of hemispheric asymmetry in response to two different colors (red and blue). The two hemispheres responded in essentially the same manner to the two colors. However, an interesting finding was the larger ERPs of both left and right hemispheres to the color blue, especially with central visual field stimulation. A third experiment was directed at the determination of whether the right hemisphere would be more sensitive to visual masking than the left, especially with a metacontrast paradigm which is dependent on spatial factors to produce the masking effect. A fourth experiment was performed to determine the relative responsivity of the two hemispheres to verbal (letter) and spatial (geometric form) stimuli.

The research sought to understand certain fundamental aspect of visual information processing in the areas of recognition, exploration and search behavior-progress was made in the following areas. Laboratory Reconfiguration. During the six months of this contract ASL undertook to upgrade and augment the eye movement recording equipment at HEL. One major effort was devoted to writing a data acquisition program to record and analyze pupil diameter data. Data Collection & Analysis. An entirely new hardware and software data acquisition environment had to be designed and built and a new I/O configuration designed for the PDP 11 computer. Keywords: Computer applications; Modeling; Optical imaging; Algorithms; Attention; Vision; Pattern recognition; Target recognition; Data storage; Laboratory equipment.

Research on several interrelated topics is described in this report. These projects are focused on the analysis of feature and conjunction detection, models of selective attention, and curve tracing. One project examines the effects of a heterogeneous background on feature search. Another assesses spatial factors (such a target-distractor separation) in the detection of targets defined in terms of simple features. A third project has the goal of developing methods for determining the extent to which processing is serial or parallel. A fourth represents initial efforts to determine whether conjoined features are represented in retinotopic or spatiotopic 'maps'. A fifth explores top-down and bottom-up factors in visual search. A sixth makes use of an inhibitory priming method to test early-and late-selection models of selective attention. Finally, a project is reported in which the operation of visual curve tracing is studied. Keywords: Visual perception; Attention; Perception; Information processing; Vision; Visual search; Curve tracing.

The publications describe progress in two related areas of visual information processing: motion processing and visual attention. The full equivalence between Reichart motion detection and Fourier motion analysis (first-order motion processing) was proved formally. A new experimental paradigm was developed to test the model of nonFourier (2nd-order) motion processing. This model, which accounts for the perception of motion-from-texture, consists of a stage of linear spatio-temporal filtering followed by fullwave rectification and then by standard (Reichart) motion analysis. It was demonstrated that human 2nd-order motion is, for practical purposes, one-dimensional (i.e., a single channel system). The spatial filter that this channel utilizes was measured and found to be lowpass. Work on attentional processes in visual task using rapid sequences of superimposed patterns showed that highly trained subjects were unable to use gross physical differences to filter out unattended items at an early stage of perceptual processing. On the contrary, the results are explained by postulating that attended and unattended elements of the input are tagged as such at an early-stage, and are then discriminated later on the of this tag. (Author)

Eye movements determine the location and velocity of the retinal image. Thus, to understand how we see it is necessary to understand both how eye movements are controlled and how they effect visual information processing. The proposed research is concerned with both problems. Specifically: the effect of expectations on smooth eye movements; and the eye moves smoothly in the direction of expected future target motion. Experiments will determine how expectations and guesses about the direction of future motion are formulated and the relative contributions of expectations and retinal image motion to smooth eye movements. The effect of saccades and saccade-like stimulus perturbations on visual information processing: Saccades continually displace the retinal image, yet we see the world as a single coherent picture. Experiments will find out whether the visual system selectively tolerates rapid lateral displacements, or whether the decision to move the eye is required. Programming sequences of saccades: Experiments will show whether sequences of saccades can be pre-programmed, and whether use of such sequences improves performance of visual tasks.

This final progress report summaries the results of a study that successfully determined the functional architecture of visual motion perception, in the sense of defining the mechanisms involved and the relations between them. It was proved that visual motion is computed by two neural systems: primitive motion-energy extraction (e.g., Reichardt detector) and higher-level feature tracking. A psychophysical pedestal paradigm was used to exclude the feature- tracking process and thereby to obtain pure measures of motion-energy extraction. Motion energy extraction was found to be exclusively monocular, fast (cutoff frequency is 12 Hz) and sensitive (can utilize 0.2% contrast), 'bottom- up', and, to operate on both luminance (first-order) and contrast (second-order) motion stimuli. Motion feature tracking was found to operate interocularly as well as monocularly, have a cutoff frequency of 3 Hz, and to be both bottom up (it computes motion from luminance, contrast, depth, motion-motion, flicker and other type of stimuli) and top-down (e.g., attentional states influence what appears to move). The full report is appended.

This final progress report summaries the results of a study that successfully determined the functional architecture of visual motion perception, in the sense of defining the mechanisms involved and the relations between them. It was proved that visual motion is computed by two neural systems: primitive motion-energy extraction (e.g., Reichardt detector) and higher-level feature tracking. A psychophysical pedestal paradigm was used to exclude the feature- tracking process and thereby to obtain pure measures of motion-energy extraction. Motion energy extraction was found to be exclusively monocular, fast (cutoff frequency is 12 Hz) and sensitive (can utilize 0.2% contrast), 'bottom- up', and, to operate on both luminance (first-order) and contrast (second-order) motion stimuli. Motion feature tracking was found to operate interocularly as well as monocularly, have a cutoff frequency of 3 Hz, and to be both bottom up (it computes motion from luminance, contrast, depth, motion-motion, flicker and other type of stimuli) and top-down (e.g., attentional states influence what appears to move). The full report is appended.

Twelve (12) experiments are described in this report. The first nine (9) are concerned with the hypothesis that the identification of the values of stimulus features in multielement visual displays requires serial processing. Contrary to this hypothesis, the weight of the evidence suggests that feature identification can be carried out by spatially parallel processes. The remaining three (3) experiments are concerned with the ability to extract semantic information from several stimuli in parallel. Both alphanumeric character classification and lexical (i.e., word vs. nonword) decisions can be accomplished by parallel processes, but semantic categorization of words cannot. The implications of these findings for theories of attention are discussed. Keywords: Attention, Lexical access, Perception, Semantic processing, Subsidizing, Texture perception, Vision, Visual search.

In man-machine design, it is important to quantify the reaction time components instead of simply determining the lump-sum reaction time to stimulus. The primary purpose of this thesis was to investigate the reaction time components, such as visual perception and muscle response time, and to quantify them by separating from their aggregated sum. The prime example, traffic-signal cognition simulation was used to examine human reaction time to signal change. With a modified computer program that stimulates the driver's approach to the intersection, we measured the subject's reaction times and examined behavioral patterns. Twelve subjects were involved in the experiment. A logistic regression procedure was applied to the data to define subjects' choices at different distances. Decision process time and the conflicting decision area were examined. Logistic regression was used to reveal the distribution of the conflicting decision area and muscle response time. The results revealed the visual perception time distribution. The most important part of total reaction time was visual perception. overall, the study showed the possibility of quantifying the reaction time components by using a simple computer simulation.

The research explored varios aspects of visual processing in the perception of features and objects. Studies of visual features tested the level of abstraction of the coding of orientation, the emergence of features in surface media defined by texture, motion or depth, and the role of similarity in preattentive representation of coarsely coded values. Object perception was explored in a series of studies requiring subjects to conjoin features, and testing the role played by spatially selective attention. The results suggested a modification of feature integration theory, allowing some preattentive control through feature maps. We also found evidence for object-specific tokens in the coding of dynamic displays of moving, changing objects. In studies of perceptual learning, extended practice effects were shown to be very specific to details of the practice situation. Finally, visual memory for arbitrary line patterns showed both surprising plasticity )long-lasting traces after a single presentation of 50 patterns) and surprising dissociations between implicit and explicit tests. (js)

The objective is to find out how saccadic eye movements (voluntary jumps of the eye) can be directed accurately to chosen targets within natural, patterned visual environments. An effective saccadic scanning mechanism must consist of: (1) a selective filter that determines which object serves as the saccadic target, and (2) a spatial-pooling mechanism that computes a single, precise saccadic landing position. Major findings were: (1) a single attentional filer serves both saccadic eye movements and perception; (2) attentional requirements of saccades are modest; (3) it is easier to divide attention between two widely-spaced targets than two closely-spaced targets; (4) the spatial-pooling process guides saccades to a central landing position within spatially-extended targets with a high degree of precision; (5) the central landing position is near the center of gravity; (6) the central landing position and the precision of the saccade depends on target size but not on spatial frequency content, contrast, or shape.

Continuing work on the three-systems-theory of visual motion perception: (I) Isoluminant chromatic motion is perceived by and only by the third-order motion system. (2) New illusion. Cancelling salience modulation in the third-order motion system causes brightly colored moving red-green gratings to appear to be motionless (motion standstill illusion) that gives considerable insight into the mechanisms of object perception. (3) A prediction that motion could be used to provide amplification of apparent contrast was verified, with amplifications 5x. (4) Amplification procedures were used to construct purified stimuli to stimulate each motion system individually, and (5) to measure the amplification produced by selective attention to a feature. (6) Developed computational models oi attentional amplification of features and of the movement dynamics of spatial attention. The spatial attentional mode was investigated with a new paradigm (gating spatial attention) to enable measurements of iconic memory uncontaminated by attention dynamics and to demonstrate that a new attention window' open concurrently at all points within the window. (7) The spatial attention model predicts 90% of the variance in a huge data set (approximately 500 points per observer from 20,000 trials) and encompasses the major attention paradigms. (All items plus more are fully described in publications.)

This investigation estimates the magnitude of the relationship between social class and visual information processing rates, and compares the development curves of processing rate for advantaged and disadvantaged children. Subjects were grade school children in a metropolitan area of southwestern Michigan. Estimates of the information processing rate were obtained by tachistocopically presenting two stimuli to the subject. The interval between the stimuli required by the subject to process the first stimulus was that subject's information processing rate. Social class was quantified using Duncan's "Socioeconomic Index". Data analysis involved use of the Pearson correlation, multiple regression technique and analysis of variance. Results indicated that disadvantaged children do process visual information more slowly than advantaged children. The development curves for the two groups tend to converge as they proceed through school. (Author/CJ)

The first year of the grant was spent in setting up the laboratory, and in starting research on a number of different projects. All are concerned with the visual processing of information in the perception of objects. A series of experiments has explored the perception of conjunctions of features, attempting to determine what makes this difficult or easy. A new method (detection of apparent motion) was tested and a modification of feature-integration theory was developed to accommodate the new results. Other projects have been concerned with coding of features, finding evidence for modularity, testing the level of abstraction at which features (such as orientation) are coded, the different media which support the coding of shape, and the space in which they are represented (retinal or three-dimensional). Another project has probed the effects of perceptual learning with extended practice at detecting particular sets of targets; the results suggest that automatization in search is highly specific to the practiced task and has little effect on other perceptual tests. Six graduate students are at present, working on projects wholly or partly supported by the grant.