Perceptual continuation and depth in visual phantoms can be explained by perceptual transparency.

We try to explain perceptual continuation and depth in the visual-phantom illusion in terms of perceptual transparency. Perceptual continuation of inducing gratings across the occluder in stationary phantoms could be explained with unique transparency, a notion proposed by Anderson (1997 Perception 26 419-453). This view is consistent with a number of previous reports including that of McCourt (1994 Vision Research 34 1609-1617) who criticized the stationary phantom illusion from the viewpoint of his counterphase lightness induction or grating induction, which might involve invalid transparency. Here we confirm that the photopic phantom illusion (Kitaoka et al, 1999 Perception 28 825-834) really gives in-phase lightness induction and involves bistable transparency. It is thus suggested that perceptual continuation and depth in the visual-phantom illusion depend on perceptual transparency.

Stationary phantoms and grating induction with oblique inducing gratings: implications for different mechanisms underlying the two phenomena.

The visibility of stationary visual phantoms and the grating induction (GI) effect were concurrently analyzed with both black and gray inspection areas (IA) using the same subjects with counterbalanced orders of measurements. Oblique inducing gratings were employed in order to compare the visibility of obliquely aligned and vertically misaligned appearances between the two phenomena. Aligned and misaligned phantom responses with a black IA were similar, whereas overall phantom visibility was severely suppressed when the IA was gray. In contrast, misaligned GI dominated with a gray IA, whereas aligned and misaligned GI responses were similar with a black IA. Phantoms appear to be related to visual mechanisms' selectively utilizing relative luminance information between the inducing grating and IA in a manner consistent with more global figural characteristics of the display (e.g., modal and amodal completion). On the other hand, GI may be predominantly due to locally operating brightness/contrast mechanisms.

Connectivity perception of partly occluded gratings in 4-month-old infants.

Four groups of eight 4-month-old infants were each habituated to one of four displays consisting of a grating of either low (0.4 cycle deg(-1) or high (1.2 cycles deg(-1) spatial frequency, whose central portion was covered up with a horizontal occluder which was either narrow (1.33 deg) or broad (4.17 deg). Posthabituation displays consisted of a complete grating of the same frequency as the habituated grating, along with a separate grating whose central portion was replaced with a black gap of the same height as the occluder in the habituation displays. All the infants, except those who were habituated to the high frequency with the broad occluder, looked longer at the separate grating than the complete grating display during posthabituation trials. Previously, we found that infants under 1 month of age perceive the grating continuation only when the occluder height is less than about 0.5 cycle of the grating; our present results show that this figure increases to about 1.6 cycles of the grating frequency in the case of 4-month-old infants. These findings indicate that those developmental changes depend on both the sufficiency of visual information available and the efficiency of the perceptual ability of infants for grasping spatial relationships.

Dynamic properties influence the perception of facial expressions.

Two experiments were conducted to investigate the role played by dynamic information in identifying facial expressions of emotion. Dynamic expression sequences were created by generating and displaying morph sequences which changed the face from neutral to a peak expression in different numbers of intervening intermediate stages, to create fast (6 frames), medium (26 frames), and slow (101 frames) sequences. In experiment 1, participants were asked to describe what the person shown in each sequence was feeling. Sadness was more accurately identified when slow sequences were shown. Happiness, and to some extent surprise, was better from faster sequences, while anger was most accurately detected from the sequences of medium pace. In experiment 2 we used an intensity-rating task and static images as well as dynamic ones to examine whether effects were due to total time of the displays or to the speed of sequence. Accuracies of expression judgments were derived from the rated intensities and the results were similar to those of experiment 1 for angry and sad expressions (surprised and happy were close to ceiling). Moreover, the effect of display time was found only for dynamic expressions and not for static ones, suggesting that it was speed, not time, which was responsible for these effects. These results suggest that representations of basic expressions of emotion encode information about dynamic as well as static properties.

Two competing mechanisms underlying neon color spreading, visual phantoms and grating induction.

Neon color spreading is closely related to the photopic visual phantom illusion, since these two completion phenomena are characterized by in-phase lightness induction, and the only difference in the stimulus configuration is the difference in the inducer height. This idea was supported by the present study. Neon color spreading showed almost the same function of critical spatial frequency as photopic visual phantoms (Experiment 1), and the critical spatial frequency was constant as the inducer height was changed (Experiment 2). We also examined the relationship between neon color spreading and grating induction (characterized by counterphase lightness induction) in critical spatial frequency (Experiment 3) and in magnitudes of lightness induction (Experiment 4) as a function of the inducer height. The inducer height at which in-phase (neon color spreading) appearance gave way to counterphase (grating) induction was approximately 0.1 deg. These results suggest that neon color spreading shares a common neural mechanism with the photopic visual phantom illusion and that this mechanism is different from, and competes with, the mechanism of grating induction.

Similarity between Petter's effect and visual phantoms.

Here we draw attention to similarity between Petter's effect and the visual phantom illusion. Phantoms are visible when the spatial frequency of the inducing grating is low or the occluder is thin, whereas phantoms are invisible when the spatial frequency of the inducing grating is high or the occluder is thick. Moreover, phantoms are perceived in front of the occluder when they are visible, whereas the occluder is seen in front of the inducing gratings when phantoms are invisible. These characteristics correspond to Petter's effect, in which the thicker region tends to be perceived in front of the thinner region when two regions of the same lightness and of different sizes overlap, since 'thick' corresponds to low spatial frequency of the inducing grating or a thick occluder while 'thin' corresponds to high spatial frequency of the inducing grating or a thin occluder.

Photopic visual phantom illusion: its common and unique characteristics as a completion effect.

An illusion similar to the stationary visual phantom illusion presented earlier by Gyoba (1983, Vision Research 23 205-211) is reported. This illusion is visible in photopic vision and we have tentatively named it the 'photopic phantom illusion'. A typical example of this illusion is a white and light-gray square-wave grating occluded by a black region. In this figure, a phantom grating running across the occluder with clear contours but less contrast, is seen. The critical spatial frequencies of photopic phantoms have been measured and compared with those of scotopic phantoms that have been reported previously, revealing a great resemblance between them. We discuss the characteristics of this illusion in terms of transparency, stereoscopic viewing, and perceptual completion.

Visual completion of partly occluded grating in infants under 1 month of age.

Four groups of eight infants (3 weeks of age on average) were each habituated to one of four displays consisting of a grating of either low (0.4 cpd) or high (1.2 cpd) spatial frequency, whose central portion was covered up with a horizontal occluder which was either narrow (1.33 degrees) or broad (4.17 degrees). These habituation displays are referred to as LN (low spatial frequency grating and narrow occluder), LB (low and broad), HN (high and narrow), and HB (high and broad) displays. Posthabituation-test displays consisted of a complete grating (CG) of the same frequency as the habituated grating along with a separate grating (SG) whose central portion was replaced with a black gap of the same height as the occluder in the habituation displays. Infants habituated to the LN display looked significantly longer at the SG than the CG display during posthabituation-test trials. Infants habituated to the LB and HN displays looked at the CG and SG displays, almost equally. In contrast, infants habituated to the HB display looked longer at the CG than the SG display. These results show that infants under 1 month of age can perceive the continuation of the grating behind the occluder, and that their visual completion on habituation displays can be evoked according to the interaction between the spatial frequency of the grating and the occluder height.

[Delays in recognizing facial expressions resulting from prolonged viewing of adaptation face stimuli].

The main purpose of the present study was to investigate the mechanism underlying fundamental dimensions in recognition of facial expressions, using the prolonged viewing method. Pictures of female faces were used, which displayed one of the following six: happiness, sadness, surprise, anger, disgust, and neutral expressions. In Experiment 1, the mean ratings of each expression on each picture were analyzed with principal component analysis (PCA). In Experiment 2, subjects orally judged the expressions of test faces following either one or 25 seconds of viewing of an adaptation face. Significant delays occurred in the prolonged viewing condition only when the adaptation face had a higher absolute value on 'pleasantness' the first component of PCA, than the test face. In contrast, adaptation faces that were high only on 'arousal,' the second component, produced no such effects. It was suggested that there are at least two subsystems involved in the recognition of facial expressions and that each system has different temporal characteristics.

Loss of a forest: perceptual fading and filling-in of static texture patterns.

Fading and filling-in of static texture patterns is demonstrated with the use of a matrix of Chinese characters with vertical or oblique line components. An asymmetric effect is demonstrated by exchanging foreground and background characters.

[Delays produced by prolonged viewing in the recognition of Kanji characters: analysis of the "Gestaltzerfall" phenomenon].

It is a well-known observation that when a Kanji character is viewed steadily and continuously, the viewer often becomes unable to recognize the Kanji as a whole pattern and it becomes difficult to judge whether or not the Kanji is orthographically correct. Such a phenomenon is called the "Gestaltzerfall" of Kanji characters. In the present study, two experiments were carried out to examine delays in the recognition of test Kanji following 25 s of prolonged viewing of adaptation Kanji, which were comprised of either the same or different parts and structures. When the size of the adaptation Kanji was equal to that of the test Kanji, there were significant delays of more than 50 ms, both when the stimuli were of the same pattern, and when they had the same structure but different components. However, when their sizes were different, delays were found only when the test and the adaptation Kanji were of the same pattern. These results suggest that a Kanji pattern may be internally represented as a whole, independent of its size, while the processing of Kanji structure may be dependent upon its size. Prolonged viewing may produce an adaptation effect specific to such representations.

Disappearance of stationary visual phantoms under high luminant or equiluminant inducing gratings.

The visibility of stationary phantoms was measured by human observers in two experiments. The phantom visibility declined with the increase of the mean luminance of black/white inducing gratings, falling near to zero at a mean luminance high above 80 cd/m2. The phantoms also disappeared when the red/green inducing gratings were subjectively equiluminant. The implications of these results are discussed in relation to the grating induction effect and to the functional correlations with the magnocellular system's activity.

Optimal occluder luminance for seeing stationary visual phantoms.

If the central portion of a vertical grating is covered up by an opaque horizontal occluder, a phantom grating is perceived to continue across the occluded region. The phantoms can be seen even if the inducing grating is stationary, but are not seen when the occluder luminance is near to the space-average luminance of the grating. Optimal occluder luminance for the perception of phantoms was measured, while occluder height and the inducing grating contrast varied. Phantom visibility was maximal when the occluder luminance was near to either the minimum or to the maximum luminance of the grating. When the occluder luminance was set between the two levels, opposite-phase illusory gratings were dominantly observed. The reciprocal relationship between the visual phantoms and the grating induction effect of McCourt (1982) is discussed.

[Reaction time and eye movements in the recognition task of hand-written Katakana-letters: an experimental verification of the discriminant analysis of letter recognition by Hayashi's Quantification].

Hayashi's Quantification, model II, is a multivariate discriminant analysis for qualitative data. This method was applied to the quantitative indices about confusability of recognition of handwritten Japanese Katakana-letters. The experimental results of reaction time and eye movements in the recognition task corresponded well with indices computed by this statistical procedure. (a) The labels, given by the subjects to the letters, were in line with the prediction by the discriminant, (b) the reaction time and the number of fixations were bigger for highly confusable letters, and (c) those features, which were important according to the discriminant, were fixated more frequently. Thus, Hayashi's quantification procedure is valid for psychological experiments.