Observable Response Research Articles

Optimal methods for calculating classification images: weighted sums.

In signal detection theory, an observer's responses are often modeled as being based on a decision variable obtained by cross-correlating the stimulus with a template, possibly after corruption by external and internal noise. The response classification method estimates an observer's template by measuring the influence of each pixel of external noise on the observer's responses. A map that shows the influence of each pixel is called a classification image. Other authors have shown how to calculate classification images from external noise fields, but the optimal calculation has never been determined, and the quality of the resulting classification images has never been evaluated. Here we derive the optimal weighted sum of noise fields for calculating classification images in several experimental designs, and we derive the signal-to-noise ratio (SNR) of the resulting classification images. Using the expressions for the SNR, we show how to choose experimental parameters, such as the observer's performance level and the external noise power, to obtain classification images with a high SNR. We discuss two-alternative identification experiments in which the stimulus is presented at one or more contrast levels, in which each stimulus is presented twice so that we can estimate the power of the internal noise from the consistency of the observer's responses, and in which the observer rates the confidence of his responses. We illustrate these methods in a series of contrast increment detection experiments.

Effects of a limited class of nonlinearities on estimates of relative weights.

Perturbation analyses have been applied in recent years to determine the relative contribution of individual stimulus components in detection and discrimination tasks. Responses to stimulus samples are compared to stimulus parameters to determine the details of the decision rule. Often, a linear model is assumed and it is of interest to determine the relative contribution of different stimulus elements to the decision. Here, biases in estimated relative weights are considered for the case where the decision variable is given by D = (sigma(alphaiXn(i))k)m and the stimulus components, the Xi, are normally distributed, of equal variance, and mutually independent. The alphai are the "true" combination weights, and n, k, and m are positive reals. The method used to estimate relative weights is the correlation coefficient between the Xi and the observer's responses. Estimates of relative alphai do not depend on m but may depend on the mean values of the Xi and the values of n and k (a dependence on the variance, sigmai2, holds even for linear transformations).

Toward a Psychophysics of Perceptual Organization Using Multistable Stimuli and Phenomenal Reports

We explore experimental methods used to study the phenomena of perceptual organization, first studied by the Gestalt psychologists. We describe an application of traditional psychophysics to perceptual organization and offer alternative methods. Among these, we distinguish two approaches that use multistable stimuli: (1) phenomenological psychophysics, in which the observer's response is assumed to accurately and directly reflect perceptual experience; and (2) the interference paradigm, in which an observer's response is evaluated as correct or incorrect because it pertains to a corrigible task, but does not directly reflect the observer's experience. We show that phenomenological psychophysics can yield valuable information about perceptual organization and lends itself to the development of quantitative theory. We discuss some criticisms of the method and argue that the two approaches that use multistable stimuli are complementary. We also compare each of the approaches with traditional psychophysics. We conclude that the several methods are convergent.

Recently reported sightings of ball lightning: observations collected by correspondence and Russian and Ukrainian sightings

The observations in the first section of this paper were sent to John Abrahamson in response to the publication of a recent paper on ball lightning, with the correspondents either reading the original paper, or reports of it in popular science articles. A selection of the cases has been made, including those which showed interesting detail possibly useful in debating ball lightning mechanisms. Any inserted text within parenthesis is the observer's response to follow-up queries from J.A. The age of the observer (where noted) is that at the time of the observation. Three observations (1 q-s) did not show motion independent of their surroundings, but have been included because of their other similarities to ball lightning. It is interesting to note the high proportion (greater than 0.5) of scientifically or technically trained observers in this collection. The data presented in the second section of this paper come from both letters and interviews. Our interview questionnaire consisted of 46 questions and was mostly carried out in quiet conditions. Observations in 2 a, e, h, l, m, q-t were corrected during several (two or three) interviews. Heading each observation case we note the most unusual property of the object. Ball lightning appears in Russia and the Ukraine usually in summer (June-August), and more rarely in spring (March-May), or autumn (September-November). It appears usually during the daytime, 13:00-17:00 h, when most summer thunderstorms take place. Observers in their descriptions usually use the term "morning" to describe the period from 06:00 to 11:00 h, "daytime" for 12:00-18:00 h, and "evening" for 19:00-21:00 h. The term "time' means the local time, and sometimes it is difficult to compare it with the local geographic time due to frequent official state summer-winter time changes. The decree time in Russia can in general differ from the geographic time by 1-2 h.

What Is Ecological Validity? A Dimensional Analysis.

Ecological validity has typically been taken to refer to whether or not one can generalize from observed behavior in the laboratory to natural behavior in the world. Although common in current discussions of research, the idea of ecological validity has a long history in psychological thought. A brief historical examination of this idea reveals that concerns with ecological validity are evident in multiple dimensions of experimental work, including the nature of the experimental setting, the stimuli under investigation, and the observer's response employed as the measure. One problem with this multidimensionality, however, is that no explicit criteria have been offered for applying this concept to an evaluation of research. One consequence of this problem is that concerns with ecological validity can be raised in most experimental situations. This article includes a discussion of some demands of ecological validity and the nature of these different dimensions, as well as a critical evaluation of research on the development of mobility with respect to these constraints.

Categorical Color Mapping Using Color-Categorical Weighting Method—Part II: Experiment

This article is linked to the previous article, “Categorical color mapping using color-categorical weighting method Part I: Theory” (Ref. 1) and shows its performance for gamut mapping through psychophysical experimentation. This method maps an original color to a reproduced color by weighting constraints mapping so as to match their color names between the original color and the reproduced color. A degree of the weighting is derived from a categorical color naming experimentation. Therefore, categorical color mapping incorporates an observer's response in calculating an optimal mapping point. The algorithm of categorical color mapping has seventeen constraints. Nine of them are given as the geometrical centers of gravity of the color categories. The others are positioned on the gamut surface. In the former case there was colorimetric matching pair-wise between a source space, and a destination space and the latter was given by categorical lightness mapping with constant hue. Categorical color mapping was evaluated as the superior method among tested methods in paired-comparison experiments. GCUSP and clipping method were applied as conventional techniques to compare to categorical color mapping. From the viewpoint of categorical color mapping, performance of GCUSP can be explained. In comparison with these conventional methods, hue linearity correction in CIELAB space by categorical color mapping was confirmed.

Discriminative Effects of Compound Drug Stimuli: A Focus on Attention

Discrimination research has increasingly used compound stimuli emerging from drugs acting through multiple neurotransmitter systems or from injections of drug mixtures that approximate “street-wise” drug-taking behaviors. Accompanying this trend has been an interest in the role of cognitive factors in drug discrimination learning. Accounts of multidimensional drug stimuli have focused mainly on specific neuronal mechanisms, and have largely ignored the contribution of stimulus information to the perception of internal events or to the selection processes, heretofore called attention mechanisms, which may underlie the observer's idiosyncratic response to drug administration. It is argued here that research in drug discrimination may benefit from a more detailed consideration of the processes by which an observer interacts with the emergent stimulus properties of drug administration. Therapeutic intervention initiatives may critically depend on knowing the interactions between the specific attributes of the drug experience that capture the attention of the individual and that may later acquire stimulus control over complex drug-taking behaviors.

A perceptually tempered display for digital mammograms.

The cathode ray tube of a workstation for use with digital mammograms was calibrated with a photometer to produce an input-output characteristic curve similar to the perceptually linear curve defined by a current display standard. Then, a test pattern consisting of bars of increasing intensity containing disks of decreasing contrast was used by an observer to estimate the minimal detectable contrast (MDC) at different levels of display luminance. The MDC was modeled by a parabola. The shape of the parabola was determined by the observer's perceptual responses, and the range was determined by the maximum and minimum pixel values of the breast parenchyma. As each mammogram was displayed, the contour of the breast was automatically found and pixels within the breast image were sampled to determine the pixel values that were used to compute the maximum and minimum pixel values. The parabola was integrated to determine the look-up table for the initial MDC-tempered display of the mammogram. Preliminary observer performance tests showed no significant differences in the accuracy and speed of three radiologists who read a set of mammograms when the MDC-tempered display was compared with the perceptually linear display.

Representing contrast detection as an eigenvalue problem.

Contrast detection can be formulated as an eigenvalue problem. One of the simplest resulting models has only two parameters. The model is space variant and employs the Hermite functions as eigenfunctions. Computing the response to a sinusoidal acuity grating yields the observer's contrast response. The model itself, however, is developed within an abstract mathematical framework which is general enough to include Fourier analysis as a special case. Consequently, the methods of Fourier analysis are generalized to those of eigenfunction expansion and the spectral theory of linear operators.

Red-green chromatic discrimination with variegated and homogeneous stimuli

Chromatic discrimination thresholds were measured under conditions which yielded fine and degraded discrimination steps. Discrimination was assessed by identification of the location of one of four homogeneous equiluminant stimuli arranged in a square or with pseudoisochromatic (PIC) figures using the stimulus design of Regan, Reffin and Mollon (Vis Res 1994; 34: 1279–1299). Stimuli were presented on CRT monitors and specified in units of cone trolands. They were viewed within a surround metameric to the equal energy spectrum. L troland threshold versus retinal illuminance (TVR) functions were measured by four-alternative spatial forced-choice staircase procedures for (1) a four 1×1° equiluminant stimuli arranged in a square and (2) ‘C’ shaped pseudoisochromatic figures in which the observer had to identify the gap location. The ‘C’ was constructed of spatially discrete patches of varying size and luminance to ensure that the observer's responses depended on chromatic signals. The TVR functions appeared V-shaped and were similar for the two paradigms. The minimum occurred near the L excitation of equal-energy white. The PIC stimuli yielded poorer discrimination with the TVR function being displaced by ∼0.5 log unit. Discrimination for stimuli degraded by luminance and spatial noise presented within an achromatic appearing surround is sharpest near the chromaticity metameric to the equal energy spectrum.

DISCRIMINATIVE EFFECTS OF COMPOUND DRUG STIMULI

Discrimination research has increasingly used compound stimuli emerging from drugs acting through multiple neurotransmitter systems or from injections of drug mixtures that approximate street-wise drug-taking behaviors. Accompanying this trend has been an interest in the role of cognitive factors in drug discrimination learning. Accounts of multidimensional drug stimuli have focused mainly on specific neuronal mechanisms, and have largely ignored the contribution of stimulus information to the perception of internal events or to the selection processes, heretofore called attention mechanisms, which may underlie the observer's idiosyncratic response to drug administration. It is argued here that research in drug discrimination may benefit from a more detailed consideration of the processes by which an observer interacts with the emergent stimulus properties of drug administration. Therapeutic intervention initiatives may critically depend on knowing the interactions between the specific attributes of the drug experience that capture the attention of the individual and that may later acquire stimulus control over complex drug-taking behaviors.

Effects of stimulus information and number of stimuli on sequential dependencies in absolute identification.

Abstract Three absolute identification experiments were conducted to examine the effects on sequential dependencies of two factors, namely stimulus information and the number of stimuli used. The data were analysed by multivariate information analysis (Mori, 1991; Mori & Ward, 1995). The results showed that the sequential dependencies became larger as the stimulus information was decreased (Experiments 1 and 3) and as the number of stimuli was increased (Experiments 1 and 2), consistent with previous studies (Gamer, 1953; McGill, 1957; Mori & Ward, 1995). These results are discussed in the context of models of sequential dependencies (e.g. Mori, 1989; Ward & Lockhead, 1971). In addition, multiple regression analysis was performed on the present data, and the results were compared with those of the information analysis. In an absolute identification task, an observer is presented with a set of stimuli and asked to identify the stimuli with a predetermined set of response categories. In sensation and perception research, the absolute identification task is of great use in measuring sensitivity to more than two stimuli (e.g., Braida & Durlach, 1972, 1988). Whereas the measures of sensitivity (e.g., d information transmission) only reflect the effect of a presented stimulus (Sn) on an observer's response (Rn), it is now well known that Rn is also affected by the immediately preceding stimulus (Sn-1) and response (Rn-1). Such effects of Sn- I and Rn- 1 are called sequential dependencies, and whether the dependencies are positive (assimilative) or negative (contrast) depends on the experimental condition (Mori, 1988; Ward & Lockhead, 1971). There is growing recognition that sequential dependencies on Sn-1 and Rn-I should be studied thoroughly in order to fully understand the observer's performance in the absolute identification task (e.g., Laming, 1984; Mori, 1988, 1989; Mori & Ward, 1995; Treisman, 1985). Previous research has shown that there are at least three factors affecting sequential dependencies in absolute identification: Stimulus information available to the observer, the number of stimuli used, and the presence or absence of trial-by-trial feedback (e.g., Gamer, 1953; McGill, 1957; Mori, 1989; Mori & Ward, 1990; Ward & Lockhead, 1971). The present paper concerns the first two factors, stimulus information available and number of stimuli. The effects of trial-by-trial feedback have been discussed extensively in another paper (Mori & Ward, 1995). STIMULUS INFORMATION AVAILABLE TO AN OBSERVER In the absolute identification task (and also other psychophysical tasks), the observer attempts to obtain information from a presented stimulus, Sn, to make a precise judgment of the stimulus. In applications of information theory to empirical research, information available from Sn to the observer is measured by information transmission from Sn to Rn (Garner, 1962). Numerous experiments have shown that the amount of information transmitted varies with changing stimulus duration (Ward, 1991), stimulus range or spacing (Braida & Durlach, 1972; Luce, Green, & Weber, 1976), viewing or listening conditions (Lockhead, 1966; Mori, 1989; Mori & Ward,1990, 1992; Ward, 1979), and presence or absence of trial-by-trial feedback (Eriksen, 1958; Mori, 1989). Several findings suggest that such a change in stimulus information also affects sequential dependencies. It has been shown that when stimulus information is decreased, the dependencies on Sn-1 and Rn-1 increase (e.g., McGill, 1957; Mori, 1989; Ward & Lockhead, 1971). For example, Mori (1989) systematically varied stimulus information by changing the contrast level of the stimuli for absolute identification of line lengths. Multivariate information analysis (Garner & McGill, 1956; McGill, 1954, 1957) was applied to the data to compute the information measures of dependency on Sn- 1 and on Rn- 1. The results showed that the dependencies on both Sn-1 and Rn-l increased as the stimulus information decreased. …

Luminance contrast and spatial-frequency tuning of the transient-vergence system.

Vergence has transient components that are stimulated by brief presentations of stimuli at large disparities (up to several degrees). The question that we have addressed is what stimulus features are encoded by this system. A competition paradigm [Jones & Kerr, (1972)]. Vision Research, 12, 1425-1430) was used in which three gabors were presented. A single Gabor was presented to the fovea of one eye and two gabors, 2.5 deg to either side of the fovea, to the other; one of which, when paired with the single Gabor defined a convergent direction, the other a divergent direction. First we determined if increasing the luminance contrast of the Gabor pair whose disparity was opposite to the observer's response-bias direction (variable-contrast pair) relative to the remaining Gabor (reference) could alter the observer's response direction. Secondly, we determined if the contrast required for such a change in response was affected by the relative spatial frequency of the convergent and divergent Gabors. The reference Gabor was held at 2 cpd and the variable Gabor pair was varied between 5.6 and 0 (a gaussian) cpd. Results demonstrated that increasing the luminance contrast of the variable pair relative to the reference Gabor could alter the observer's response direction, even when the contrast of only one of the variable-pair Gabors was increased. The luminance contrast required for this change to occur was directly related to the spatial frequency of the variable pair over the entire frequency range tested. Vergence responses were preferentially made to lower spatial frequencies, even when a low spatial frequency was pared with a high one. We conclude that transient-vergence responses are not reduced by mixed contrasts (i.e. no contrast-paradox effect) and appear to be mediated by a system that employs a single lowpass sensitive channel.

Countertransference in psychoanalytic psychotherapy with children and adolescents

This is a rich, compact volume. Despite the seminal article by Winnicott, “Hate and the Counter-transference” (1949), child therapists have been slow to discuss countertransference issues. Children's preference for action, their closeness to primary process thinking, and the role of the therapist as play participant interact to enhance countertransference reactions. Preoedipal preverbal traumata communicated in nonverbal and regressive modes tend to evoke preoedipal issues in the therapist that may have been poorly or incompletely analyzed. Nevertheless, countertransference can become a resource in working with the child, particularly with regard to those split-off parts that become “contained” in the therapist. The countertransference issues in child analysis and psychotherapy entail emotional interactions with and by parents who, while supportive of the treatment and its goals, cannot tolerate the very changes so fervently desired. In the introduction, John Tsiantis's brief summaries of each of the 11 chapters serve to create bridges and connections between them. Anastopoulous and Tsiantis provide an exhaustive review of countertransference issues, achieving an excellent balance of completeness, condensation, and historical viewpoints enhanced by their own critical comments. This chapter could easily become required reading in the training of all child psychotherapists. In a brief chapter, Judith Trowell demonstrates how the skills required to observe young children, if focused on the observer's inner responses, provide exceptionally useful information about the inner child. Alex Holder spent 20 years at the Anna Freud Center treating children four to five times per week and the last 10 years in Hamburg, where government restrictions limit sessions to twice a week. The significant differences in containment, regression, momentum, intensity, elaboration of fantasies, and depth of transference are well addressed. Holder emphasizes that all of these specific patient factors are mirrored in the analyst as well. The clear advantages of high-frequency meetings are identified. Holder's unique experiences and his thoughtful and incisive comments are more worthy of repeated reading than summarization. Holder conceptualizes the differences convincingly. Anne-Marie Sandler discusses the countertransference difficulties of the therapist when the patient is overwhelmed by the shame of a worthless and devalued self-image and similarly refuses to value the help and the person of the therapist. In this way the patient is spared the intolerable disappointment of an object relationship. Francois Ladame takes the theoretical position that language is the “unique tool” of psychoanalysis, and cautions us neither to “despise” nor to “prize” countertransference. Ladame redefines countertransference as a challenge from the patient that is experienced by the therapist as “narcissistically frightening.” Ladame suggests group discussions rather than individual supervision in helping the therapist deal with these anxieties. Using clinical vignettes, Jacqueline Godfrind shows how readily countertransference errors can arise from the parents' attitudes toward the child, toward the treatment's progress, and toward the therapist herself. Didier Houzel uses a Kleinian orientation of “psychic envelopes,” part/whole objects, and bi-sexuality involved in the primitive splitting of maternal and paternal part objects. I found the material hard to comprehend and the clinical vignettes unconvincing. Tsiantis discusses the formidable challenge of an inpatient setting for traumatized adolescents. Challenges in this setting include splitting in the transference, projective identifications with the staff, the balance between the therapeutic individual role and the administrative team role, and the multiple opportunities for acting out of transference/countertransference roles among all involved. Tsiantis attributes success to the training and the “stamina” of the nursing staff, their consultation, and their treatment as respected team members. The traumatized adolescent in long-term residence is adept at recreating with the staff his or her own traumatic circumstances. Tsiantis demonstrates how experience, thoughtfulness, and attention to the subtle and not so subtle interplay between transference and countertransference among all those involved can be therapeutic. Ten pages of valuable references are complemented by a very serviceable index. There might be some difficulties for the American reader in encountering terms and conceptualizations that are not in everyday use here, but it is obvious that the authors know more about analytic work in the United States than we know about the Europeans. I think the book will be used as a standard text (perhaps also the advanced text) for those therapists who treat children and their parents in depth. I suspect it will become the standard reference and a classic in its time.

Deriving instrumental tolerances from pass-fail and colorimetric data

Equations such as CIE94 and CMC are now in common use to set instrumental tolerances for industrial color control. A visual experiment was performed to generate a data set to be used in evaluating typical industrial practices. Twenty-two observers performed a pass-fail color tolerance experiment for a single high-chroma yellow color. Thirty-two glossy samples varying in all three CIE-LAB dimensions were compared with a single standard. A near-neutral anchor pair was used to define the quality of match criterion. The pooled pass data were used to fit a 95% confidence ellipsoid. The chromaticness dimension was well estimated by either CMC or CIE94. The lightness dimension was poorly estimated by either equation. Evaluating the sampling distribution of the 32 test samples via a covariance matrix revealed a poor sampling, particularly in the ΔL*Δb* plane. This sampling may have biased the visual experiment. The visual data were used to optimize various color-difference equations based on CIE94 and CMC, where the l:c and total color difference were adjustable parameters. Several methods of optimization are described including minimizing the number of instrumental wrong decisions and logistic multiple-linear regression. Some methods require only pass response data, while others require both pass and fail data. Because industrial tolerances are usually based on a single observer, ellipsoids were fitted for three observers to demonstrate the large variability between observers in judging color differences. It was concluded that when tolerances need to be set based on a single observer's visual responses of samples not well distributed about the standard, typical industrial occurences, one should only adjust the tolerance magnitude based on a statistically valid equation such as CIE94. One should not change l:c or derive a new ellipsoid. © 1996 John Wiley & Sons, Inc.

Perceptual Classification Images from Vernier Acuity Masked by Noise

Letting external noise rather than internal noise limit discrimination performance allows information to be extracted about the observer's stimulus classification rule. A perceptual classification image is the correlation over trials between the noise amplitude at a spatial location and the observer's responses. If, for example, the observer followed the rule of the ideal observer, the response correlation image would be an estimate of the ideal observer filter, the difference between the two unmasked images being discriminated. Perceptual classification images were estimated for a Vernier discrimination task. The display screen had 48 pixels deg−1 horizontally and vertically. The no-offset image had a dark horizontal line of 4 pixels, a 1 pixel space, and 4 more dark pixels. Classification images were based on 1600 discrimination trials with the line contrast adjusted to keep the error rate near 25%. In the offset image, the second line was one pixel higher. Unlike the ideal observer filter (a horizontal dipole), the observer perceptual classification images are strongly oriented. Fourier transforms of the classification images had a peak amplitude near 1 cycle deg−1 and an orientation near 25 deg. The spatial spread is much more than image blur predicts, and probably indicates the spatial position uncertainty in the task.

The robustness of infants' detection of visual motion

Fourteen-week-olds were tested for detection of the movement of a small vertical bar with varying levels of visual complexity using a speeded version of the forced-choice preferential looking (FPL) paradigm. The target bar horizontally oscillating at temporal frequencies of 0.6, 1.2 or 2.4 Hz could appear amidst 1, 5, or 13 other static bars. Detection accuracy was (a) generally unaffected by the complexity of the visual field, (b) affected by the distance of the moving bar from the center of the screen, and (c) affected directly by temporal frequency. Response times depended inversely on temporal frequency. When the similarity of the static elements was decreased by making them chromatically heterogeneous, the FPL observer's average response time increased, although detection accuracy remained the same. The FPL observer's response times actually decreased slightly as more heterogeneous distractors were added. Adults tested using an FPL observer also showed this same effect. Both infants and adults showed disproportionately long response times on incorrect trials when only 1 other distractor was present in the visual field. The number of static elements in the visual field and their similarity affected primarily the speed with which infants at this age responded to movement.

Effects of retinal disparity and screen format on three-dimensional object position and course-prediction judgments

A human factors study in which participants viewed 3-D imagery under two task conditions designed to mimic the critical information processing elements of an aircraft control application is reported. In one task, observers judged the relative distance between objects in a 3-D workspace, whereas the other task required observers to extrapolate object positions from course heading cues. Operator performance was assessed for imagery presented with or without retinal disparity cues in three display formats (i.e. , plan view, simple perspective view, and enhanced perspective view) and at three signal-to-clutter levels. The findings indicate that retinal disparity cues reduce the number and magnitude of response errors in the course prediction task, but do not affect observer's response times or ratings of response confidence. The effects of display format varied between the two tasks. In the relative distance assessment task, response times for the simple perspective format were shorter than those for the plan view or enhanced perspective formats. In the course prediction task, response times for the plan view and simple perspective formats were shorter than for the enhanced perspective format. Display format did not affect response errors, error magnitude, or ratings of response confidence.

On the time course and accuracy of spatial localization: Basic data and a two-process model

This article addresses the question how fast and accurate the location of a single stimulus can be perceived. In Experiment 1, we measured localization performance in a task which required subjects to perceive and report the location of a single target stimulus (‘ *’ sign) presented in one square of an imaginary 25×19 grid. Two factors were varied: stimulus duration and stimulus eccentricity. Stimulus duration was manipulated by employing a backward masking stimulus. Ten intervals (stimulus onset asynchronies) separated target and masking stimulus: 25, 50, 75, 100, 125, 150, 200, 250, 300, and 350 ms. Stimulus eccentricity was manipulated by presenting the target stimulus at five different distances from the fixation point. The observer localized the target stimulus by moving the cursor from the middle of the grid (the initial fixation point) to the perceived target location by pressing the ‘arrow’ keys on the keyboard. Localization performance showed to be typically related to stimulus duration. That is, two components could be distinguished: The first component represented an initial steep rise in localization performance during the first 50 ms of stimulus duration; the second component represented a gradual rise in localization performance after 50 ms, reaching maximal performance at about 300 ms. We interpreted these two localization performance functions as reflecting the operation of two systems, namely the attentional system for the initial strong increase and the eye movement system for the subsequent gradual increase. In Experiment 2, we measured saccadic eye response latencies to clarify the role of eye movements in localization performance. It was found that in 98.4% of all trials saccades were executed, and, moreover, that saccadic eye response latency decreased with increasing stimulus duration. In Experiment 3, we compared localization performance in the absence and presence of eye movements and demonstrated that localization performance for stimulus durations up to 50 ms was independent of eye movements. Overall, the present findings were interpreted as evidence in support of a two-process model of localization performance in which a shift of attention is followed by a rapid eye movement toward the target location. In line with a continuous flow conception of visual information processing, our model assumes that location information takes time to develop in the visual system; hence, an observer's localization response may be based on qualitatively different processes operating on qualitatively different kinds of information. In case of short duration stimuli, information conveyed by transient cells is used by the attentional system to shift attention toward the target location; this results in course location information being available. In case of longer duration stimuli, information conveyed by sustained cells is used by the eye movement system to prepare and generate a saccade toward the target location. Fine location information therefore depends on the accuracy of the generated saccade which in turn depends on the duration of the stimulus. A final gaze hypothesis is advanced to capture the importance of eye position information in manual aiming performance. Finally, it is emphasized that the present findings require confirmation and extensions; possible directions for future research are discussed.

Application of the EM algorithm to radiographic images.

The expectation maximization (EM) algorithm has received considerable attention in the area of positron emitted tomography (PET) as a restoration and reconstruction technique. In this paper, the restoration capabilities of the EM algorithm when applied to radiographic images is investigated. This application does not involve reconstruction. The performance of the EM algorithm is quantitatively evaluated using a "perceived" signal-to-noise ratio (SNR) as the image quality metric. This perceived SNR is based on statistical decision theory and includes both the observer's visual response function and a noise component internal to the eye-brain system. For a variety of processing parameters, the relative SNR (ratio of the processed SNR to the original SNR) is calculated and used as a metric to compare quantitatively the effects of the EM algorithm with two other image enhancement techniques: global contrast enhancement (windowing) and unsharp mask filtering. The results suggest that the EM algorithm's performance is superior when compared to unsharp mask filtering and global contrast enhancement for radiographic images which contain objects smaller than 4 mm.