Matching Stimuli Research Articles

Use of image-based information in judgments of surface-reflectance properties.

We examined how well we can recover surface-reflectance properties from shading patterns under changes in surface shape. The stimulus we used was a square surface modulated in depth by a low-pass-filtered random field and rendered by the Phong illumination model [Commun. ACM 18, 311 (1975)]. Two different surface images (target and match) were presented side by side, with either the viewing direction or the surface-normal direction rotating around the horizontal axis. The target shape was manipulated by changing the spatial spectrum, and the target reflectance was manipulated by changing the diffuse-reflection coefficient and the specular-reflection exponent (shininess) of the Phong model. The shape parameters of the match stimulus were fixed, but its reflectance parameters were under the control of subjects, who had to make the apparent reflectance of the two surfaces as similar as possible. The results showed that the constant error (difference between simulated and matched values) was large except when the two surfaces had the same shape parameters or when they differed only in scale. The pattern of the constant errors and response variabilities suggests that the judgments of the subjects were based on the similarity of the luminance histogram of the surface image. Our results demonstrate a limitation of surface-reflectance constancy for changes in shape and the importance of image-based information in reflectance judgments. The results are discussed in relation to previous studies that showed effects of spatial layout on surface-reflectance perception.

Color appearance: neutral surrounds and spatial contrast

The experimental data in this paper show that chromatic bars presented in alternation with equiluminant neutral-appearing bars are seen as more saturated than the same chromaticity presented as a uniform rectangle. This effect was diminished but not eliminated when test and match stimuli were presented within a slightly dimmer neutral surround. The test stimulus was a 2×5° rectangle with a 0.7 cpd square wave grating composed of alternating equiluminant chromatic test bars and neutral bars. Asymmetric matching was used to match the test bar appearance to a uniform 2×5° comparison rectangle. Test and comparison stimuli were presented to separate eyes in a haploscope and appeared flanking a central fixation target. Data were collected with test and inducing stimuli on the cardinal axes of color space. Test bars separated by neutral bars appeared more saturated than the comparison rectangle for both the l-and s-axis directions. Manipulation of excitation on one cardinal axis did not affect the appearance matches made for the other cardinal axis.

Event-related potentials during digit recognition tasks

Objective : An event-related potential (ERP) correlate of visual short-term memory (VMP) has been identified in our laboratory. This study aims to determine how stimulus load modulates recognition processing of digits. Methods : ERPs were recorded from 117 healthy right-handed subjects during a delayed matching-to-sample paradigm, using number stimuli that were either low load (three digits long) or high load (five digits long). The bootstrap method [R. Srebro, A bootstrap method to compare the shapes of two scalp fields, Electroenceph. Clin. Neurophysiol. 100 (1996) 25–32.] was employed to evaluate the topographic features of the VMP revealed in the current source density (CSD) maps. Results : Response times were significantly shorter for matching stimuli than for non-matching stimuli only for low loads; longer response times were related to higher loads compared to low loads only for matching stimuli. The high loads were related to larger ERP responses. The ERP was significantly smaller for matching than for non-matching three-digit numbers, but not for five-digit numbers. The ERP was also reduced to the test stimuli compared to sample stimuli regardless of stimulus load. Both temporal and frontal regions were involved in the recognition of the digit stimuli, and the left hemisphere was more active in the non-matching processing of digits. Conclusions : The VMP spatial pattern in addition to its amplitude is sensitive to stimulus load in the encoding process.

Adaptive shift of visual sensitivity balance under ambient illuminant change.

We examined the relationship between the ambient illuminant chromaticity and changes in the sensitivity balance of the visual system, using illuminants of various chromaticities. The sensitivity of observers was measured in a room with a variable-chromaticity illuminant. The observer's state of chromatic adaptation was measured with unique-white settings. Our results showed that the change in visual sensitivity has a nonlinear correlation with the change in illuminant chromaticity; chromatic adaptation was nearly complete for desaturated illuminants, but the degree of chromatic adaptation became worse as the illuminant became more saturated. We defined a new index, relative cone weights, which represents this relationship well. To measure the role of chromatic induction from the immediate-surround area of the matching stimulus, we performed additional experiments by presenting the test inside a colored or black immediate surround. The results showed that the unique-white settings were not disturbed by the change in immediate-surround color. Our results imply that the room illuminant chromaticity was the primary factor in changing the observer's state of chromatic adaptation.

Colour memory matching analysed using different representation spaces

The choice of the representation space is relevant when analysing the results of a colour memory experiment. If we are interested in determining which colours are best or worst remembered using the colour difference between the reference and matching stimuli, we need spaces with a high degree of uniformity. In this paper, the results of a previous experiment with 10 colours, 100 observers and three time gaps have been processed in five representation spaces with different uniformity degrees: CIE , CIE , SVF, Rlab and Llab. For most colours and in all the spaces, reference colours above a particular lightness value are remembered as even lighter, whereas below a certain value they are remembered as darker or equally light. Chroma either increases or suffers no variations. Changes in hue appear to be, in contrast, unsystematic. These results basically agree with the previous literature. CIE , CIE and Rlab spaces yield practically identical results. The best remembered colours are orange and oxide red, and yellow is the worst remembered colour. In the SVF space, the worst remembered colours are blue and yellow, although the results for pink become comparable as the delay increases. Pink is also the worst remembered colour in the Llab space.

Integrating electrophysiological and anatomical experimental data to create a large-scale model that simulates a delayed match-to-sample human brain imaging study.

We propose a model that draws together experimental evidence from anatomical, electrophysiological and imaging experiments in order to understand better the neural substrate of human imaging studies using positron electron tomography (PET) and functional magnetic resonance imaging (fMRI). First, we define a simple local circuit that reflects the major role that local connectivity plays in producing PET and fMRI data, which are thought to mainly reflect synaptic activity. Second, in order to account for the role of varying behaviors during the course of a typical imaging experiment, we propose a local circuit that can perform a delayed match-to-sample task. The elements of this circuit behave very much like neurons that have been found in the prefrontal cortex during similar tasks in monkeys. One subpopulation responds selectively only when stimuli are present. Two different populations show the two types of delay-period activity that have been identified, one with high activity both during the cue and the delay period, the other with a rise during the delay period only. Last, a subpopulation shows a brief response only if the second stimulus matches the first, thus mediating the decision about whether the stimuli match. We show that in addition to performing the task, the integrated summed synaptic activities of the model are similar to experimental PET data.

The role of utricular stimulation in determining perceived postural roll-tilt

We measured perceived postural orientation in roll in 5 subjects in two conditions which generate perception of bodily roll-tilt: firstly during roll-tilt on a tilt-chair in a 1-g environment, and secondly during constant velocity rotation 1 metre from the axis of rotation on a fixed-chair human centrifuge. Perceived postural orientation was measured by the subjects' setting of a small motor-driven bar of light-emitting diodes, rotatable in the frontal plane, to the perceived gravitational vertical. The stimulus values on the tilt-chair and centrifuge were selected so that they generated the same interaural utricular shear to the otolithic receptors in the inner ear. For such matched utricular shear stimuli, the subjects showed significantly smaller roll-tilt perception on the centrifuge than on the tilt-chair. This result shows that it is not utricular shear alone which is important for roll-tilt perception. Instead, at the stimulus values used here, perception corresponds to the angle of the resultant force, which is not given by utricular shear alone but from the ratio of utricular to saccular input. Somatosensory input may complement this calculation.

Reflection of working memory: ERP mnemonic effects

The study of working memory often utilizes a delayed matching to sample paradigm (DMS). Typically in the matching condition, the test and sample stimuli are identical, raising the possible confound of retinotopic projections for the matching stimuli in contrast to the non-matching stimuli. In the present study, 65 healthy subjects performed a modified delayed matching to sample task while monitoring their ERP waveforms. The stimuli consisted of 60 different sample stimuli (S1) and 60 different test stimuli (S2). Half of the S2s were complementary to the sample stimuli (Fit), the other half of the S2s were not complementary (Nonfit). After S2, the subjects pressed one of the buttons to indicate whether the test stimulus fits the sample stimulus. Our statistical results indicated that the ERPs to sample stimuli differed from the ERPs to test stimuli from 200 ms poststimulus to the end of the recording epoch. The ERPs to fitting stimuli were significantly different from those to non-fitting stimuli from 200 to 400 ms poststimulus. The ERP patterns in the present study may reflect ERP mnemonic effect for working memory. Our results ruled out the retinotopic confound as a potential mediator variable, and are in agreement with other animal or human neurophysiological studies on memory.

Is working memory intact in alcoholics? An ERP study

Few investigators have applied the working memory theory to studies on abstinent chronic alcoholics, though it has been reported that the deficits in short-term memory appear to be specific to visuo-spatial and problem-solving tasks. In the present study, we recorded ERPs from 40 male control subjects and 78 alcoholics performing a modified delayed matching to sample task. To minimize the possible confound of retinotopic projections for the matching stimuli, in contrast to the non-matching stimuli, we employed a unique set of stimuli in our delayed matching to sample task. Our results indicate that an ERP component, occurring at approximately 250 ms post-stimulus, may be a reflection of the ERP mnemonic effect for working memory. This component distinguishes the two groups at the right occipitotemporal region, providing evidence of right hemisphere dysfunction in alcoholics. Thus, the current experiment may show electrophysiological evidence of working memory deficits in alcoholics.

A Quantitative Method to Explore the Assimilation Effects

The role of central processes in the assimilation effect can easily be shown qualitatively (de Weert and Spillmann, 1995 Vision Research35 1413 – 1419), but it is difficult to measure quantitatively because of the subtlety of the effect. In most experimental designs, the match stimulus differs greatly in appearance from the test stimulus, eg in size or configuration, and because these differences are far more striking than the assimilation effect, matching is difficult. Central processing, eg object segmentation, influences colour spreading. It is this property that we explored with a new approach: a matching task in which the match stimulus has the same properties (eg size and configuration) as the test stimulus. Object segmentation is forced by stereopsis-induced depth. The test stimulus consists of two depth planes, one with black dots and the other with white dots, on a homogeneous gray background. The match stimulus has the same configuration of black and white dots, but now squeezed into a single depth plane. The basic idea behind this stimulus is that assimilation mainly acts on the back plane of a scene (as can be shown experimentally). So, while keeping the appearance of the stimulus the same, subjects can focus on the assimilation effect itself. This new approach allows us to explore more aspects of the assimilation effect and gain insight into the processes involved.

Object and place memory in the macaque entorhinal cortex.

Lesions of the entorhinal cortex in humans, monkeys, and rats impair memory for a variety of kinds of information, including memory for objects and places. To begin to understand the contribution of entorhinal cells to different forms of memory, responses of entorhinal cells were recorded as monkeys performed either an object or place memory task. The object memory task was a variation of delayed matching to sample. A sample picture was presented at the start of the trial, followed by a variable sequence of zero to four test pictures, ending with a repetition of the sample (i.e., a match). The place memory task was a variation of delayed matching to place. In this task, a cue stimulus was presented at a variable sequence of one to four "places" on a computer screen, ending with a repetition of one of the previously shown places (i.e., a match). For both tasks, the animals were rewarded for releasing a bar to the match. To solve these tasks, the monkey must 1) discriminate the stimuli, 2) maintain a memory of the appropriate stimuli during the course of the trial, and 3) evaluate whether a test stimulus matches previously presented stimuli. The responses of entorhinal cortex neurons were consistent with a role in all three of these processes in both tasks. We found that 47% and 55% of the visually responsive entorhinal cells responded selectively to the different objects or places presented during the object or place task, respectively. Similar to previous findings in prefrontal but not perirhinal cortex on the object task, some entorhinal cells had sample-specific delay activity that was maintained throughout all of the delay intervals in the sequence. For the place task, some cells had location-specific maintained activity in the delay immediately following a specific cue location. In addition, 59% and 22% of the visually responsive cells recorded during the object and place task, respectively, responded differently to the test stimuli according to whether they were matching or non-matching to the stimuli held in memory. Responses of some cells were enhanced to matching stimuli, whereas others were suppressed. This suppression or enhancement typically occurred well before the animals' behavioral response, suggesting that this information could be used to perform the task. These results indicate that entorhinal cells receive sensory information about both objects and spatial locations and that their activity carries information about objects and locations held in short-term memory.

Analysis of phantom source localization cues

A simple and widely used method for controlling the azimuthal direction of an apparent ‘‘phantom’’ sound source between two frontal loudspeakers is to vary the level difference between the (otherwise identical) signals delivered to the speakers. Explanations of phantom source localization [e.g., D. M. Leakey, J. Acoust. Soc. Am. 31, 977–986 (1959); B. B. Bauer, J. Acoust. Soc. Am. 33, 1536–1539 (1961)] have been based on the fact that, at low frequencies, the interaural phase delay in the dual-source stimulus matches that of an actual single source from some direction within the angular bounds of the loudspeakers. However, this theory fails to account for very similar localization results with stimuli that contain only high-frequency energy. It is also inconsistent with the perceptual robustness of the phantom image, which is unitary even with wideband stimuli. Leakey (1959) proposed that an interaural group delay underlies the high-frequency phantom. The present analysis of stimulus cues suggests that the interaural level difference may be an equally potent cue at high frequencies. Taken as a whole, this analysis points up the remarkable finding that this simple phantom-image technique results in the important interaural differences being well-matched to those of a single actual source across the audio spectrum. [Work supported by NIH.]

The postimperative negative variation following ambiguous matching of auditory stimuli

Previous studies using the delayed-matching-to-sample paradigm with visual stimuli reliably induced a postimperative negative variation (PINV) in both schizophrenic patients and healthy controls. The PINV was found to be: (a) larger in schizophrenic patients than controls; (b) larger under conditions of ambiguous as compared to clear matching conditions; and (c) larger over right-hemispheric, fronto-temporal regions in controls, while a less asymmetrical distribution with a tendency for left-frontal predominance occurred in schizophrenic patients. The present study examined to what extent the development and scalp distribution of the PINV were modality-specific by applying the delayed-matching-ro-sample design using auditory stimuli. Furthermore, the neurophysiological state during anticipatory negativity (CNV) and PINV was examined by presenting acoustic probe stimuli (one per trial on 50% of the trials) during baseline, CNV or PINV interval. Event-related slow and probe-evoked potentials were recorded in 13 patients with a chronic schizophrenic disorder (DSM-III-R) and 13 healthy control Ss from 15 electrode sites including midline and two sagittal rows over each hemisphere. Comparable group differences and effects of ambiguity on PINV amplitudes were found for both modalities, visual and auditory. The auditory stimuli produced a fronto-central distribution of the PINV in both groups. The probe-evoked vertexpotential was smaller in patients compared to controls, but exhibited comparable modulation with the largest amplitudeduring the CNV in both groups. Results suggest modality-non-specific cognitive determinants of the PINV. However, stimulus modality did affect the scalp distribution of the PINV. Probe-evoked responses point to different functional meaning of negativities prior to (CNV) and following (PINV) task-relevant stimuli. © 1997 Elsevier Science B.V.

Transposition errors in visual matching of orthographic stimuli: A study of normal children with implications for orton's theory of developmental dyslexia

Transposition (or sequence) errors, such as reading DAM as MAD, have been frequently reported in dyslexic subjects and in (younger) normal readers. Nonetheless, the theoretical status of these errors has remained controversial. We report a visual matching experiment, with Roman and Cyrillic letter-strings, in which one of the foils contains the letters of the stimulus in reversed order. Errors in which this foil is chosen are very common in preschool Italian children (ages 4–6 years), but reduce precipitously after only a few months of learning to read (in a school setting). Normal children, it seems, rapidly learn that concatenations of letters are not commutative in modern orthographies. Furthermore, in the first and second grade school children, there is no significant correlation between the number of errors they make on either of the visual discrimination tasks and their performance in reading aloud.

Limitations of surface-color and apparent-color constancy.

Color-constancy mechanisms have been studied and discussed in a number of investigations. However, there has been little attempt to reveal how color constancy deteriorates as the conditions for it become less than optimal. We carried out a series of asymmetric color-matching experiments, using two criteria: surface-color match and apparent-color match. With brief adaptation the degree of color constancy increased as chromatic cues were added in the surround. In the condition of black surround, the test stimuli appeared self-luminous, and chromaticities of the chosen matching stimuli were the same as the physical chromaticities of the test stimulus, indicating a total deficiency of color constancy. With 15 min of preadaptation to the illuminant, the surface-color matches showed almost perfect color constancy under illuminant change. In both adaptation conditions, the chromatic-shift of matches from what would be expected for perfect color constancy increased gradually between 1,700- and 30,000-K illuminant, as chromaticity of the illuminant departed from 6,500-K illuminant. Under 1,000-K illuminant the surface-color appearance became totally achromatic, and color constancy was completely lost. Our results show that, even with brief adaptation to the illuminant, the contribution of the surrounding stimulus is large enough to achieve a fair degree of color constancy, but complete adaptation to the illuminant helps to achieve almost perfect color constancy.

Flicker brightness enhancement and visual nonlinearity

The purpose of this study was to investigate the nonlinear mechanism underlying brightness enhancement, in which a flickering stimulus appears brighter than a steady stimulus of equal mean luminance. The flickering and matching stimuli were temporally alternated. Both were cosine windowed to minimize the potential effects of temporal transients. Subjects adjusted the amplitude of the matching stimulus to match it in brightness to the flickering stimulus. The temporal frequency, modulation, and waveform of the flickering stimulus were varied. With sinusoidal flicker, brightness enhancement increased with increasing modulation at all frequencies, peaking at about 16 Hz at full modulation. The results were modeled by a broad temporal filter followed by a single accelerating nonlinearity. The derived temporal sensitivity of the early filter inferred from brightness enhancement decreased more slowly at high frequencies than the filter(s) inferred from flicker modulation thresholds. With low frequency sawtooth flicker, brightness enhancement was phase-dependent at low, but not at high modulations, suggesting that multiple neural mechanisms may also be involved in addition to an early nonlinearity.

High resolution evoked potential imaging of the cortical dynamics of human working memory

High resolution evoked potentials (EPs), sampled from 115 channels and spatially sharpened with the finite element deblurring method, were recorded from 8 subjects during working memory (WM) and control tasks. The tasks required matching each stimulus with a preceding stimulus on either verbal or spatial attributes. All stimuli elicited a central P200 potential that was larger in the spatial tasks than in the verbal tasks, and larger in the WM tasks than in the control tasks. Frequent, non-matching stimuli elicited a frontal, positive peak at 305 msec that was larger in the spatial WM task relative to the other tasks. Irrespective of whether subjects attended to verbal or spatial stimulus attributes, non-matching stimuli in the WM tasks also elicited an enhanced P450 potential over the left frontal cortex, followed by a sustained potential over the superior parietal cortex. A posterior P390 potential elicited by infrequent, matching stimuli was smaller in amplitude for both spatial and verbal WM tasks compared to control tasks, as was a central prestimulus CNV. These results indicate that WM is a function of a distributed system with both task-specific and task-independent components. Lesion studies and coarse temporal resolution functional imaging methods, such as PET and fMRI, tend to paint a fairly static picture of the cortical regions which participate in the performance of WM tasks. In contrast, the fine-grain time resolution provided by imaging brain function with EP methods provides a dynamic, picture of subsecond changes in the spatial distribution of WM effects over the course of individual trials, as well as evidence for differences in the activity elicited by matching and non-matching stimuli within sequences of trials. This information about the temporal dynamics of WM provides a critical complement to the fine-grain spatial resolution provided by other imaging modalities.

The effects of multiple‐talker stimuli on immediate memory span.

Previous experiments revealed that changing the voice between study items on an auditory list of letters results in a decrease in memory span [H. M. Saldaña, J. Acoust. Soc. Am. (1995)]. This finding is contrary to Baddeley’s articulatory loop model which defines memory span in terms of time or duration of items. According to this model items are stored as memory traces that fade after approximately 2 s unless revived by an articulatory control process [A. D. Baddeley and G. J. Hitch, Psychol. Learn. Motiv. (1974)]. The number of items that can be reactivated within the decay time can be retained indefinitely. The present study replicates and extends the previous memory span findings by utilizing digits instead of letters and by using a stimulus matching methodology for estimating span length. A reconceptualization of the articulatory loop model is discussed which takes into account the encoding of stimulus variability. [Work supported by NIH.]

Temporal constraints on the McGurk effect.

Three experiments are reported on the influence of different timing relations on the McGurk effect. In the first experiment, it is shown that strict temporal synchrony between auditory and visual speech stimuli is not required for the McGurk effect. Subjects were strongly influenced by the visual stimuli when the auditory stimuli lagged the visual stimuli by as much as 180 msec. In addition, a stronger McGurk effect was found when the visual and auditory vowels matched. In the second experiment, we paired auditory and visual speech stimuli produced under different speaking conditions (fast, normal, clear). The results showed that the manipulations in both the visual and auditory speaking conditions independently influenced perception. In addition, there was a small but reliable tendency for the better matched stimuli to elicit more McGurk responses than unmatched conditions. In the third experiment, we combined auditory and visual stimuli produced under different speaking conditions (fast, clear) and delayed the acoustics with respect to the visual stimuli. The subjects showed the same pattern of results as in the second experiment. Finally, the delay did not cause different patterns of results for the different audiovisual speaking style combinations. The results suggest that perceivers may be sensitive to the concordance of the time-varying aspects of speech but they do not require temporal coincidence of that information.

Sources of Disagreement in Object Naming

Two experiments examined the processing of objects with low name agreement. Experiment I compared naming latencies for objects with three different types of name disagreement to those for matched control objects with very high name agreement. Objects with low name agreement due to abbreviations (e.g. phone) were named no more slowly than were control objects. Objects with multiple names (e.g. couch, sofa, settee) and objects often given incorrect names (e.g. spider for ant) took longer to name correctly than did matched controls. These results were confirmed in a second naming experiment using a revised set of high-name-agreement control stimuli. In Experiment 2, subjects carried out an object decision task using the revised stimulus set. Subjects could recognize objects with multiple names as quickly as those with high name agreement. Objects often given incorrect names were recognized by subjects more slowly than were high-agreement matched stimuli. The pattern of data suggests that the delay in naming latency due to the availability of more than one correct name arises after structural recognition. In contrast, the slowed naming of objects often misnamed would seem to originate from difficulties encountered at or before the structural stage of recognition.