Colour perception

Opposing Effects of Colorations added to Humorous Illustrations

Valeria Biasi, Sara Longo, Luna Fiorani, Daria D'Aloise and Paolo Bonaiuto

A series of works formed the basis for creating a new classification of colorations divided into reassuring and playful hues as compared to alarming and serious ones [Biasi, Bonaiuto and Giannini, 2007, in: Colore e Colorimetria: Contributi Multidisciplinari Vol. III., A. Rizzi, Florence, Centro Ed. Toscano]. The reassuring and playful colours are pink, orange, sky blue, pale yellow, light green and other pastel hues, while the alarming and serious colours include black, grey, purple, olive green and dark blue. A large number of humorous illustrations on various topics and published in specific journals were first collected. For each topic, three typical illustrations were then chosen and arranged in various versions: black and white; in acrylics for the alarming and serious version or for the reassuring and playful one; and a mixed version. Each version was evaluated individually by adults of both genders who them gave a humour score ranging from 0 (minimum) to 10 (maximum). It was hypothesised that humour scores would be significantly higher with the reassuring and playful colours, lower with the alarming and serious hues, and intermediate with the mixed or black-and-white versions. The results constantly confirmed the working hypotheses.

Dynamic desaturation illusion

Kohske Takahashi, Shun'Ya Yamada, Fuminori Ono and Katsumi Watanabe

When windmill patterns (like “*”) and disks (like “o”) filled with the same color are displayed alternately, observers fail to notice that the windmill patterns are colored; they see uniformly desaturated windmill patterns and colored disks alternating. This perceptual desaturation has the following characteristics: (1) A brighter background induces stronger desaturation; the strength depends almost linearly on the luminance of the background. (2) Darker windmills produce stronger desaturation. (3) The desaturation occurs only when the corresponding pair of windmill pattern and their masking disk is in the same color. (4) The desaturation occurs only when the pair of windmill pattern and their masking disk occupy the same retinal positions. (5) The perceptual desaturation occurs rapidly (within a few second). (6) The apparent desaturation occurs almost simultaneously for all the windmill patterns presented in the visual field. Since the windmills and masks must be filled with the same color and be at the same retinal position, we conjecture that a rapid adaptation process might enhance and interact with simultaneous contrast in this strong and quick perceptual desaturation.

Edges and gradients in real 3-D color scenes

John Mccann and Alessandro Rizzi

Color appearance can be analyzed either in terms of an object's reflectance and scene illumination, or in terms of edges and gradients. In one, appearance should correlate with reflectance because humans discount the spectra and non-uniformity of illumination. In the other, both reflectance and illumination make equal contributions to scene radiances. Edges generate large changes in appearance, while gradient are much less visible. Experiments using painted cubes, 3-D Mondrians, and real scenes measured the appearance of objects with constant reflectance in complex real-world illumination. We measured scene radiances either with a meter, or with special RAW image calibration [Vonikakis, https://sites.google.com/site/vonikakis/research/raw]. We matched the appearances of identical surface reflectances in variable illumination. The results show that matches of appearance correlate with the spatial configuration of edges and gradients, rather than with reflectance and discounted illumination.

Test of dichromatic simulation by visual search paradigm

Shoji Sunaga, Tomomi Ogura and Takeharu Seno

We investigated the validity of the dichromatic simulation [Brettel et al, 1997, Journal of Optical Society of America A, 14, 2647-2655] with a visual search paradigm. The stimuli consisted of 13 colored disks (one target and 12 distractors) on the background of the D65 chromaticity. Two colors were assigned to 6 disks as distractor colors. First, the reaction time required to find the target was measured. The reaction times for trichromatic observers in the dichromatic simulated color stimulus task tended to be shorter than those obtained from deuteranomalous observers in an original color stimulus task. This suggests that it is necessary to increase the difficulty of the detection of the target color in the simulated stimuli. Next, we tested a modified dichromatic simulation that incorporates a reduction of excitation purity. We obtained a quantitatively good prediction when the excitation purity was reduced by a factor of 0.36. Finally, we propose a new adaptive dichromatic simulation model based on the color difference between the simulated target color and the simulated distractor color in order to obtain a better predictor of reaction time in the visual search task for color defects.

Colour discrimination threshold determination using pseudoisochromatic test plates obtained by photographic and inkjet printing

Kaiva Luse, Maris Ozolinsh and Sergejs Fomins

Pseudoisochromatic test plates are widely used to diagnose the type and severity of colour vision deficiencies. The aim of the study is to determine best printing technology for creation of colour vision deficiency tests. Tests for protanopia and deuteranopia were created from perceived colour matching experiments for printed colour samples by colour deficient individuals. Calibrated EpsonStylus Pro 7800 printer for ink prints and Noritsu HD 3701 digital printer for photographic prints were used. Multispectral imagery (by tunable liquid crystal filters system CRI Nuance Vis 07) data analysis show that in case of ink prints, the measured pixel colour coordinate dispersion (along the x and y axis in the CIExy colour diagram) of equal colour arrays is smaller than in case of photographic printing. The printing quality in terms of colour dispersion for printing methods used is much higher than in case of commercially available colour vision deficiency tests. Currently most of the clinically applied colour vision deficiency tests (Ishihara, H.R.R.) classify colour vision deficit individuals in three groups – mild, medium or strong. Tests for finer classification of colour vision deficiencies are to be created based on psychophysical experiment setup to determine threshold of colour discrimination for each individual.

Red Color Enhances Memory Performance

Haruyuki Kojima

The effect of color on cognitive task performances has been reported that red enhances performance on a detail-oriented task where blue does on creative task (Mehta & Zhu, 2009, Science, 232, 1226-9). We replicated the effect to the variations of tasks. METHODS: Sixty-six students participated in the experiment. Participants were randomly assigned into either group of red or blue background color condition. They ran cognitive tasks, such as Memory Task, Remote Association Test, Proof-reading Task, etc. The tasks were conducted in a standard, dimly lit classroom. The stimuli and the instructions of tasks were presented in black, and projected on a screen with either background color of red or blue. The illuminance and the color coordinate of red color were 47.0 lx, (0.59, 0.37), while those of blue were 46.5 lx, (0.14, 0.06), respectively. RESULTS: Percent correct of memory task was significantly higher in red than blue condition. The results of the other cognitive tasks showed no significant difference in behavioral performance between the two color conditions. However, “creativity points” rated to the response words in Association Test and Imagination Tests, by independent evaluators, showed consistently and significantly higher for blue condition than red.

Perceptual Qualities and Material Classes

Christiane B. Wiebel, Karl R. Gegenfurtner and Roland W. Fleming

In daily life we often refer to materials based on their subjective properties (softness, roughness, colour) or as belonging to a specific class (metal, wood, fabric). Here, we studied how these two representations interact and compared visual and cognitive representations of material categories. Nine subjects rated 130 images from the FLICKR material database on nine different material properties (glossiness, transparency, colourfulness, roughness, hardness, coldness, fragility, naturalness and prettiness). Additionally, 65 subjects filled out a questionnaire in which they judged six different material categories according to 42 adjectives, based on their imagination. In both studies we found a high degree of consistency between subjects. To identify the most important perceptual dimensions we applied PCA to the ratings. Interestingly, we found all the materials could be represented in a meaningful way and were systematically clustered within a 2D space. In both experiments, one axis broadly captured the properties hard versus soft, while the other axis captured transparency and roughness. This suggests a systematic relationship between abstract perceptual qualities and material class. A procrustes analysis revealed that the PCA spaces from the two experiments were highly consistent, indicating a large accordance between the visual and cognitive representation of material categories tested here.

Perceptual tone-mapping operator based on multiresolution contrast decomposition

Xavier Otazu

Tone-mapping operators (TMO) are used to display high dynamic range (HDR) images in low dynamic range (LDR) displays. Many computational and biologically inspired approaches have been used in the literature, being many of them based on multiresolution decompositions. In this work, a simple two stage model for TMO is presented. The first stage is a novel multiresolution contrast decomposition, which is inspired in a pyramidal contrast decomposition (E.Peli, JOSA, 1990, 7(10), 2032-2040). This novel multiresolution decomposition represents the Michelson contrast of the image at different spatial scales. This multiresolution contrast representation, applied on the intensity channel of an opponent colour decomposition, is processed by a non-linear saturating model of V1 neurons (Albrecht et al., 2002, J. of Neurophysiology, 88(2), 888-913). This saturation model depends on the visual frequency, and it has been modified in order to include information from the extended Contrast Sensitivity Function (e-CSF) (Otazu et al., 2010, Journal of Vision, 10(12):5). A set of HDR images in Radiance RGBE format (from CIS HDR Photographic Survey and Greg Ward database) have been used to test the model, obtaining a set of LDR images. The resulting LDR images do not show the usual halo or color modification artifacts.

Spatial frequency overestimation with retinal eccentricity for isoluminant chromatic gratings

Matteo Toscani, Matteo Valsecchi and Karl Gegenfurtner

The human visual system overestimates the perceived spatial frequency of luminance gratings in the peripheral visual field [Thorpe et al., 1987, JOSA, 4,1606-1611]. This is thought to be due to a failure to adapt the labeling of spatial-frequency channels in the periphery, where spatial pooling causes the channel peak sensitivities to shift towards lower spatial frequencies [Thorpe, 1990, JOSA, 7,286-296]. This shift is more pronounced for isoluminant chromatic stimuli than for luminance stimuli [Mullen et al, 2005, Perception,34,951-959]. Therefore the peripheral overestimation of spatial frequency should increase more steeply with eccentricity for color than for luminance gratings. We asked the observers to adjust the spatial frequency of luminance and isoluminant chromatic peripheral gratings to match a foveally presented grating. Observers were tested at 2,4,6 and 8 degrees of visual angle. Spatial frequency overestimation increased with eccentricity with a steeper slope for isoluminant chromatic gratings than for luminance gratings, even when the contrast of the stimuli was variable. Since real-world edges typically consist of combinations of luminance and color, this should amount to a serious calibration problem for our visual system.

Effects of chromatic (red-green) and achromatic orthogonal masks on perceived contrast of luminance targets

Stephane Clery, Julie M. Harris and Marina Bloj

Masking has been used to study spatial vision and how the visual system combines oriented spatial filters. The common findings are suppressive effects; but in the context of detection tasks facilitatory effects have been shown when target and mask were of similar contrast. The aim of our experiment was to investigate whether these effects are present for supra-threshold contrast perception. Participants matched the contrast of a right oblique achromatic component of a plaid to a separate right oblique achromatic grating with the same spatial properties (0.75 cpd and 4 deg diameter). The target stimuli were set to several contrast values (0, 0.04, 0.15, 0.32, 1). We used two types of left oblique masks: chromatic (low or high contrast) and achromatic (0.02, 0.04, 0.2); plus a control condition without mask. Participants were generally veridical in their match without masks. Depending on the participant, luminance and chromatic contrast showed suppressive, enhancing or no effects when compared to the baseline condition. We show that the chromatic data cannot be explained by luminance artefacts. This emphasises the individual differences in contrast perception that should be taken into account if we are to fully understand the mechanisms underlying shape-from-shading and colour-shading effects.

Gradient processing: excitatory and inhibitory interactions between achromatic and chromatic mechanisms

Luis Garcia-Suarez and Marina Bloj

Previous studies using gratings have found inhibitory interactions (cross-masking) between the ‘independent’ chromatic and achromatic post-receptoral mechanisms [Switkes et al., 1988, JOSA A, 5(7), 1149-1162; Mullen and Losada, 1994, JOSA A, 11(12), 3136-3151; Chen et al., 2000, Vision Research, 40, 789-803]. Here, we investigated the interactions between the achromatic and isoluminant red-green mechanisms for gradient perception. Our (4?) gradient stimulus consisted of smooth variations in luminance and/or chromaticity over half period of a sinusoid (sine phase at the centre of stimulus) without repetition of cycles. Isoluminant points were determined for each observer (n=3) by heterochromatic flicker. We measured discrimination thresholds between test+pedestal and pedestal versus increased pedestal contrasts (TVC functions) for uncrossed (luminance test on luminance pedestal or chromatic test on chromatic pedestal) and crossed conditions (chromatic test on luminance pedestal or vice e versa) using a 2-AFC QUEST procedure. In our experiments, the same luminance pedestal had opposite effects on the detection of a chromatic test (masking) and luminance test (facilitation). In addition to the expected cross-masking, we also found some weak cross-facilitation. Most recent models of post-receptoral interactions do not account these effects; we are currently developing candidate models for gradient perception that do.

Effect of relative luminance and chromatic contrast and contrast polarity on hyperacuity thresholds: do luminance and chromatic information independently contribute to vernier performance?

Bonnie Cooper and Barry B. Lee

We have shown that luminance and chromatic channels both access mechanisms used in hyperacuity judgments [Sun et al.,2012, Vision Res., 56, 28-37]; we used grating pairs where luminance and/or chromatic contrast polarities were the same or opposite. Additionally, physiological results from macaque retinal ganglion cells were incorporated. Here we test if luminance or chromatic content contribute independently to hyperacuity judgments using grating pairs containing luminance and chromatic contrast that was varied independently with respect to both polarity and relative contrast. With luminance and chromatic contrast polarity in phase, hyperacuity thresholds with both luminance and chromatic contrast present corresponded to the lower threshold measured when the luminance or chromatic gratings were presented alone. This suggests independent access for luminance and chromatic channels to hyperacuity mechanisms. Consistent with this view, when contrast polarity of the chromatic component was reversed, luminance hyperacuity thresholds were unaffected. However, an out-of-phase luminance mask substantially elevated thresholds relying on chromatic contrast. These psychophysical data were also related to physiological measurements. The results suggest luminance signals are preferentially used as a hyperacuity cue. We interpret this as support for the idea that luminance and chromatic information is maintained independently in the afferent signal for cortical vision processing

The effect of smile and illumination color on age estimation from faces

Marcel Lucassen, Theo Gevers and Hamdi Dibeklioglu

We asked participants to 1) rate the emotion type and strength of human facial expressions from still images showing persons with a neutral expression and a spontaneous smile, and 2) estimate the age of the persons. These images are shown on a calibrated LCD display and are rendered under neutral (D65) illumination and under two simulated illuminants ”happy” and ”sad”. The latter two are constructed such that a perfectly white reflector takes on the same chromaticity and luminance as the average colors that are associated with faces showing happiness and sadness [Da Pos and Green-Armytage, Colour, Design and Creativity, 2007 1(1):2,1-20]. Both male and female persons are shown in the images, 84 in total, ranging in age from 8 to 76 years, not wearing glasses. Preliminary results, based on the data from 15 participants, show that neutral faces are most frequently categorized as “happiness”, followed by “sadness” (this holds for all three illuminants). On average, estimated ages are above the actual age. Below the actual age of 40, smiling makes the estimated age higher than for the neutral face, and vice versa above the actual age of 40. This effect is even slightly amplified by the “happy” and “sad” illuminants.

Color does not guide eye movements: Evidence from a gaze-contingent experiment

Hans Trukenbrod, Simon Barthelmé, Felix Wichmann and Ralf Engbert

Color plays a crucial role in everyday life and supports actions like searching specific objects. Whether this results from modified eye guidance has not systematically been explored. Using natural scenes, we investigated the influence of color on eye movements in a visual search task. The availability of color information was limited to a constant area around fixation by presenting gaze-contingent stimuli. The remaining image was masked by a luminance-matched grayscale version of the scene. Across trials, we used six different mask sizes. Fixated stimuli ranged from black-and-white to fully colored images. Before each trial, participants were instructed to look for a bullseye-shaped target defined either by luminance or luminance plus color. Our results show that color information was not used to guide eye movements. Except for minor disruptions in conditions with small masks, fixation durations and saccade amplitudes did not differ across conditions. Even looking for a specific color target did not change statistical measures of eye guidance. While it is beyond question that color supports vision, our results suggest that color does not modify eye-movement control. While color might help to facilitate processes like object segmentation, oculomotor control seems to be unaffected by color information.

S-cone contribution to non-assimilative color spreading in the watercolor configuration

Mikako Kuroki and Eiji Kimura

A colored line flanking a darker contour will appear to spread its color over the area enclosed by the line (watercolor effect). The watercolor effect has been characterized as an assimilative effect, but non-assimilative spreading has also been demonstrated with the same spatial configuration; e.g., when a black inner contour is paired with a blue outer contour, yellowish spreading can be observed (Pinna & Grossberg, 2005, JOSA, 22, 2207-2221). In investigating color spreading in the watercolor configuration, we found that the two types of color spreading could be observed simultaneously but on the opposite side of the contours. The assimilative spreading occurred in the area enclosed by the lighter contour, while the non-assimilative spreading occurred in the area enclosed by the darker contour. Furthermore, we also found that S-cone signals played a critical role in the non-assimilative spreading; the induced color generally varied along the S-cone axis in the chromaticity diagram and the magnitude of the spreading correlated with differences in S-cone contrast between the inner and outer contours. These results suggest that the interaction between luminance and S-cone signals is critical for the non-assimilative spreading, and that different visual mechanisms contribute to the assimilative and non-assimilative spreading.

Variation in the vision of a rectangular sequence, interspersed in a achromatic or monochromatic grating

Jorge Montalvá Colomer, Ignacio Tortajada Monta?ana and Mariano Aguilar Rico

Like what happens with achromatic gratings and chromatic sequences (paper presented at the X National Optics Meeting in Zaragoza, Spain) the value of the Bezold effect on the vision of the sequence depends on the orientation of the grating. With red monochromatic grating, the difference between the vertical and horizontal orientations (higher value with the horizontal orientation) is almost constant on the order of 0.02 to 0.04 increasing the angular frequency of the grating. This behavior is different from the green or blue monochromatic gratings, although the value of the Bezold effect is still greater with the horizontal orientation. The behavior of this difference blue or green monochromatic gratings with the angular frequency of the grating is different from the red monochromatic grating. In blue gratings the difference increases (although slightly) with the angular frequency of the grating for frequencies below 2 cycles/degree (approximately), decreasing for higher frequencies. In green gratings the difference between the Bezold effect of two orientations (horizontal - vertical) also decreases with lower frequencies of 2 cycles/degree (approximately), greatly increasing (0.01 to 0.10) with higher frequencies.

Perceptual misbinding of colour in rivalrous motion stimuli

Ryan T. Maloney, Sarah K. Lam and Colin W. G. Clifford

The correct and coherent binding of distinct visual features (colour, motion, orientation, etc) is a hallmark of normal vision. The binocular rivalry paradigm is useful for studying these interactions as it provides a situation in which conflicting images at corresponding retinal locations compete for perceptual dominance. Based on the existing evidence, it has been argued that chromatic information carried by parvocellular channels tends to promote rivalry, whereas motion signals carried predominantly by magnocellular channels tend to combine interocularly [He et al., 2005, in: Binocular Rivalry, D Alais and R Blake, Cambridge, MIT Press]. Somewhat counter to this, a misbinding of isoluminant colour within rivalrous form has recently been reported, such that the dominant percept consists of colours sourced from both eyes [Hong and Shevell, 2006, Visual Neuroscience, 23, 561-566]. Investigating this effect using counter-rotating isoluminant colour gratings, we found that while stimulus motion tended to rival, colour tended to misbind: the colours of both gratings appeared to be carried in the motion of a single grating. This suggests that the determinants of rivalry are not necessarily the same for colour and motion and supports the idea that motion rivalry can occur in isolation (Alais and Parker, 2006, Neuron, 52, 911-920).

Organic Light-Emitting Diode Monitors in Vision Science

Tobias Elze, Christopher Taylor and Peter Bex

Organic light-emitting diode (OLED) monitors, although currently still niche products on the display market, are becoming more and more popular. While the dominant display technology, liquid crystal displays (LCDs), has been analyzed and criticized with respect to their temporal behavior, OLED devices have not yet been assessed for applicability in vision research experiments. Here, we analyze the photometric and colorimetric output of OLEDs. In contrast to LCDs, OLED monitors work without backlights and can display darker black levels. However, we found cases where the color primaries were not able to display the full luminance range of 8 bit, which may impair experiments relying on color. Moreover, we found duty cycles per frame which were substantially less than one. The resulting high amplitude frequency component at the refresh rate can be within the window of visibility of neurons in the visual system and may impair applications in visual neuroscience like single unit recordings. On the other hand, compared to LCDs, luminance transitions on OLEDs occur almost instantaneously and are independent of preceding frames. Therefore, OLEDs are considerably more suitable than LCDs for experiments that require high temporal precision. Their temporal precision is comparable to that of cathode ray tube (CRT) monitors.

Temporal discrimination of flashing colour stimuli under chromatic background

Rytis Stanikunas, Algimantas Svegzda, Lina Saveikyte, Henrikas Vaitkevicius, Remigijus Bliumas, Ausra Daugirdiene and Vaiva Kulbokaite

The temporal sensitivity of human visual system was investigated. The two separate lights which flash at different temporal intervals were presented under coloured background. The three primary LED’s (red, green and blue) were used as flash lights. The background extending full visual field was illuminated by one of the primary LED’s or D65 illuminant composed of three primary LED’s. Our results suggest that background colour has influence on temporal discrimination threshold increasing sensitivity for some colours and reducing it for others. Also, the background colour induces change in flash colour perception for some combinations of background and stimuli colour. It seems likely, that the temporal discrimination process has three stages. At the first stage stimuli are perceived as flashing together. At the second stage flashes are perceived as different in time, but it is not possible to tell which stimulus flashes first. At the third stage it is clearly visible which stimulus flashed first. [Supported by the Research Council of Lithuania MIP-23/2010 and MIP-013/2012]

 

Opposing Effects of Colorations added to Humorous Illustrations Valeria Biasi, Sara Longo, Luna Fiorani, Daria D'Aloise and Paolo Bonaiuto A series of works formed the basis for creating a new classification of colorations divided into reassuring and playful hues as compared to alarming and serious ones [Biasi, Bonaiuto and Giannini, 2007, in: Colore e Colorimetria: Contributi Multidisciplinari Vol. III., A. Rizzi, Florence, Centro Ed. Toscano]. The reassuring and playful colours are pink, orange, sky blue, pale yellow, light green and other pastel hues, while the alarming and serious colours include black, grey, purple, olive green and dark blue. A large number of humorous illustrations on various topics and published in specific journals were first collected. For each topic, three typical illustrations were then chosen and arranged in various versions: black and white; in acrylics for the alarming and serious version or for the reassuring and playful one; and a mixed version. Each version was evaluated individually by adults of both genders who them gave a humour score ranging from 0 (minimum) to 10 (maximum). It was hypothesised that humour scores would be significantly higher with the reassuring and playful colours, lower with the alarming and serious hues, and intermediate with the mixed or black-and-white versions. The results constantly confirmed the working hypotheses.
Dynamic desaturation illusion Kohske Takahashi, Shun'Ya Yamada, Fuminori Ono and Katsumi Watanabe When windmill patterns (like “*”) and disks (like “o”) filled with the same color are displayed alternately, observers fail to notice that the windmill patterns are colored; they see uniformly desaturated windmill patterns and colored disks alternating. This perceptual desaturation has the following characteristics: (1) A brighter background induces stronger desaturation; the strength depends almost linearly on the luminance of the background. (2) Darker windmills produce stronger desaturation. (3) The desaturation occurs only when the corresponding pair of windmill pattern and their masking disk is in the same color. (4) The desaturation occurs only when the pair of windmill pattern and their masking disk occupy the same retinal positions. (5) The perceptual desaturation occurs rapidly (within a few second). (6) The apparent desaturation occurs almost simultaneously for all the windmill patterns presented in the visual field. Since the windmills and masks must be filled with the same color and be at the same retinal position, we conjecture that a rapid adaptation process might enhance and interact with simultaneous contrast in this strong and quick perceptual desaturation.
Edges and gradients in real 3-D color scenes John Mccann and Alessandro Rizzi Color appearance can be analyzed either in terms of an object's reflectance and scene illumination, or in terms of edges and gradients. In one, appearance should correlate with reflectance because humans discount the spectra and non-uniformity of illumination. In the other, both reflectance and illumination make equal contributions to scene radiances. Edges generate large changes in appearance, while gradient are much less visible. Experiments using painted cubes, 3-D Mondrians, and real scenes measured the appearance of objects with constant reflectance in complex real-world illumination. We measured scene radiances either with a meter, or with special RAW image calibration [Vonikakis, https://sites.google.com/site/vonikakis/research/raw]. We matched the appearances of identical surface reflectances in variable illumination. The results show that matches of appearance correlate with the spatial configuration of edges and gradients, rather than with reflectance and discounted illumination.
Test of dichromatic simulation by visual search paradigm Shoji Sunaga, Tomomi Ogura and Takeharu Seno We investigated the validity of the dichromatic simulation [Brettel et al, 1997, Journal of Optical Society of America A, 14, 2647-2655] with a visual search paradigm. The stimuli consisted of 13 colored disks (one target and 12 distractors) on the background of the D65 chromaticity. Two colors were assigned to 6 disks as distractor colors. First, the reaction time required to find the target was measured. The reaction times for trichromatic observers in the dichromatic simulated color stimulus task tended to be shorter than those obtained from deuteranomalous observers in an original color stimulus task. This suggests that it is necessary to increase the difficulty of the detection of the target color in the simulated stimuli. Next, we tested a modified dichromatic simulation that incorporates a reduction of excitation purity. We obtained a quantitatively good prediction when the excitation purity was reduced by a factor of 0.36. Finally, we propose a new adaptive dichromatic simulation model based on the color difference between the simulated target color and the simulated distractor color in order to obtain a better predictor of reaction time in the visual search task for color defects.
Colour discrimination threshold determination using pseudoisochromatic test plates obtained by photographic and inkjet printing Kaiva Luse, Maris Ozolinsh and Sergejs Fomins Pseudoisochromatic test plates are widely used to diagnose the type and severity of colour vision deficiencies. The aim of the study is to determine best printing technology for creation of colour vision deficiency tests. Tests for protanopia and deuteranopia were created from perceived colour matching experiments for printed colour samples by colour deficient individuals. Calibrated EpsonStylus Pro 7800 printer for ink prints and Noritsu HD 3701 digital printer for photographic prints were used. Multispectral imagery (by tunable liquid crystal filters system CRI Nuance Vis 07) data analysis show that in case of ink prints, the measured pixel colour coordinate dispersion (along the x and y axis in the CIExy colour diagram) of equal colour arrays is smaller than in case of photographic printing. The printing quality in terms of colour dispersion for printing methods used is much higher than in case of commercially available colour vision deficiency tests. Currently most of the clinically applied colour vision deficiency tests (Ishihara, H.R.R.) classify colour vision deficit individuals in three groups – mild, medium or strong. Tests for finer classification of colour vision deficiencies are to be created based on psychophysical experiment setup to determine threshold of colour discrimination for each individual.
Red Color Enhances Memory Performance. Haruyuki Kojima The effect of color on cognitive task performances has been reported that red enhances performance on a detail-oriented task where blue does on creative task (Mehta & Zhu, 2009, Science, 232, 1226-9). We replicated the effect to the variations of tasks. METHODS: Sixty-six students participated in the experiment. Participants were randomly assigned into either group of red or blue background color condition. They ran cognitive tasks, such as Memory Task, Remote Association Test, Proof-reading Task, etc. The tasks were conducted in a standard, dimly lit classroom. The stimuli and the instructions of tasks were presented in black, and projected on a screen with either background color of red or blue. The illuminance and the color coordinate of red color were 47.0 lx, (0.59, 0.37), while those of blue were 46.5 lx, (0.14, 0.06), respectively. RESULTS: Percent correct of memory task was significantly higher in red than blue condition. The results of the other cognitive tasks showed no significant difference in behavioral performance between the two color conditions. However, “creativity points” rated to the response words in Association Test and Imagination Tests, by independent evaluators, showed consistently and significantly higher for blue condition than red.
Perceptual Qualities and Material Classes Christiane B. Wiebel, Karl R. Gegenfurtner and Roland W. Fleming In daily life we often refer to materials based on their subjective properties (softness, roughness, colour) or as belonging to a specific class (metal, wood, fabric). Here, we studied how these two representations interact and compared visual and cognitive representations of material categories. Nine subjects rated 130 images from the FLICKR material database on nine different material properties (glossiness, transparency, colourfulness, roughness, hardness, coldness, fragility, naturalness and prettiness). Additionally, 65 subjects filled out a questionnaire in which they judged six different material categories according to 42 adjectives, based on their imagination. In both studies we found a high degree of consistency between subjects. To identify the most important perceptual dimensions we applied PCA to the ratings. Interestingly, we found all the materials could be represented in a meaningful way and were systematically clustered within a 2D space. In both experiments, one axis broadly captured the properties hard versus soft, while the other axis captured transparency and roughness. This suggests a systematic relationship between abstract perceptual qualities and material class. A procrustes analysis revealed that the PCA spaces from the two experiments were highly consistent, indicating a large accordance between the visual and cognitive representation of material categories tested here.
Perceptual tone-mapping operator based on multiresolution contrast decomposition Xavier Otazu Tone-mapping operators (TMO) are used to display high dynamic range (HDR) images in low dynamic range (LDR) displays. Many computational and biologically inspired approaches have been used in the literature, being many of them based on multiresolution decompositions. In this work, a simple two stage model for TMO is presented. The first stage is a novel multiresolution contrast decomposition, which is inspired in a pyramidal contrast decomposition (E.Peli, JOSA, 1990, 7(10), 2032-2040). This novel multiresolution decomposition represents the Michelson contrast of the image at different spatial scales. This multiresolution contrast representation, applied on the intensity channel of an opponent colour decomposition, is processed by a non-linear saturating model of V1 neurons (Albrecht et al., 2002, J. of Neurophysiology, 88(2), 888-913). This saturation model depends on the visual frequency, and it has been modified in order to include information from the extended Contrast Sensitivity Function (e-CSF) (Otazu et al., 2010, Journal of Vision, 10(12):5). A set of HDR images in Radiance RGBE format (from CIS HDR Photographic Survey and Greg Ward database) have been used to test the model, obtaining a set of LDR images. The resulting LDR images do not show the usual halo or color modification artifacts.
Spatial frequency overestimation with retinal eccentricity for isoluminant chromatic gratings. Matteo Toscani, Matteo Valsecchi and Karl Gegenfurtner The human visual system overestimates the perceived spatial frequency of luminance gratings in the peripheral visual field [Thorpe et al., 1987, JOSA, 4,1606-1611]. This is thought to be due to a failure to adapt the labeling of spatial-frequency channels in the periphery, where spatial pooling causes the channel peak sensitivities to shift towards lower spatial frequencies [Thorpe, 1990, JOSA, 7,286-296]. This shift is more pronounced for isoluminant chromatic stimuli than for luminance stimuli [Mullen et al, 2005, Perception,34,951-959]. Therefore the peripheral overestimation of spatial frequency should increase more steeply with eccentricity for color than for luminance gratings. We asked the observers to adjust the spatial frequency of luminance and isoluminant chromatic peripheral gratings to match a foveally presented grating. Observers were tested at 2,4,6 and 8 degrees of visual angle. Spatial frequency overestimation increased with eccentricity with a steeper slope for isoluminant chromatic gratings than for luminance gratings, even when the contrast of the stimuli was variable. Since real-world edges typically consist of combinations of luminance and color, this should amount to a serious calibration problem for our visual system.
Effects of chromatic (red-green) and achromatic orthogonal masks on perceived contrast of luminance targets. Stephane Clery, Julie M. Harris and Marina Bloj Masking has been used to study spatial vision and how the visual system combines oriented spatial filters. The common findings are suppressive effects; but in the context of detection tasks facilitatory effects have been shown when target and mask were of similar contrast. The aim of our experiment was to investigate whether these effects are present for supra-threshold contrast perception. Participants matched the contrast of a right oblique achromatic component of a plaid to a separate right oblique achromatic grating with the same spatial properties (0.75 cpd and 4 deg diameter). The target stimuli were set to several contrast values (0, 0.04, 0.15, 0.32, 1). We used two types of left oblique masks: chromatic (low or high contrast) and achromatic (0.02, 0.04, 0.2); plus a control condition without mask. Participants were generally veridical in their match without masks. Depending on the participant, luminance and chromatic contrast showed suppressive, enhancing or no effects when compared to the baseline condition. We show that the chromatic data cannot be explained by luminance artefacts. This emphasises the individual differences in contrast perception that should be taken into account if we are to fully understand the mechanisms underlying shape-from-shading and colour-shading effects.
Gradient processing: excitatory and inhibitory interactions between achromatic and chromatic mechanisms Luis Garcia-Suarez and Marina Bloj Previous studies using gratings have found inhibitory interactions (cross-masking) between the ‘independent’ chromatic and achromatic post-receptoral mechanisms [Switkes et al., 1988, JOSA A, 5(7), 1149-1162; Mullen and Losada, 1994, JOSA A, 11(12), 3136-3151; Chen et al., 2000, Vision Research, 40, 789-803]. Here, we investigated the interactions between the achromatic and isoluminant red-green mechanisms for gradient perception. Our (4º) gradient stimulus consisted of smooth variations in luminance and/or chromaticity over half period of a sinusoid (sine phase at the centre of stimulus) without repetition of cycles. Isoluminant points were determined for each observer (n=3) by heterochromatic flicker. We measured discrimination thresholds between test+pedestal and pedestal versus increased pedestal contrasts (TVC functions) for uncrossed (luminance test on luminance pedestal or chromatic test on chromatic pedestal) and crossed conditions (chromatic test on luminance pedestal or vice e versa) using a 2-AFC QUEST procedure. In our experiments, the same luminance pedestal had opposite effects on the detection of a chromatic test (masking) and luminance test (facilitation). In addition to the expected cross-masking, we also found some weak cross-facilitation. Most recent models of post-receptoral interactions do not account these effects; we are currently developing candidate models for gradient perception that do.
Effect of relative luminance and chromatic contrast and contrast polarity on hyperacuity thresholds: do luminance and chromatic information independently contribute to vernier performance? Bonnie Cooper and Barry B. Lee We have shown that luminance and chromatic channels both access mechanisms used in hyperacuity judgments [Sun et al.,2012, Vision Res., 56, 28-37]; we used grating pairs where luminance and/or chromatic contrast polarities were the same or opposite. Additionally, physiological results from macaque retinal ganglion cells were incorporated. Here we test if luminance or chromatic content contribute independently to hyperacuity judgments using grating pairs containing luminance and chromatic contrast that was varied independently with respect to both polarity and relative contrast. With luminance and chromatic contrast polarity in phase, hyperacuity thresholds with both luminance and chromatic contrast present corresponded to the lower threshold measured when the luminance or chromatic gratings were presented alone. This suggests independent access for luminance and chromatic channels to hyperacuity mechanisms. Consistent with this view, when contrast polarity of the chromatic component was reversed, luminance hyperacuity thresholds were unaffected. However, an out-of-phase luminance mask substantially elevated thresholds relying on chromatic contrast. These psychophysical data were also related to physiological measurements. The results suggest luminance signals are preferentially used as a hyperacuity cue. We interpret this as support for the idea that luminance and chromatic information is maintained independently in the afferent signal for cortical vision processing
The effect of smile and illumination color on age estimation from faces Marcel Lucassen, Theo Gevers and Hamdi Dibeklioglu We asked participants to 1) rate the emotion type and strength of human facial expressions from still images showing persons with a neutral expression and a spontaneous smile, and 2) estimate the age of the persons. These images are shown on a calibrated LCD display and are rendered under neutral (D65) illumination and under two simulated illuminants ”happy” and ”sad”. The latter two are constructed such that a perfectly white reflector takes on the same chromaticity and luminance as the average colors that are associated with faces showing happiness and sadness [Da Pos and Green-Armytage, Colour, Design and Creativity, 2007 1(1):2,1-20]. Both male and female persons are shown in the images, 84 in total, ranging in age from 8 to 76 years, not wearing glasses. Preliminary results, based on the data from 15 participants, show that neutral faces are most frequently categorized as “happiness”, followed by “sadness” (this holds for all three illuminants). On average, estimated ages are above the actual age. Below the actual age of 40, smiling makes the estimated age higher than for the neutral face, and vice versa above the actual age of 40. This effect is even slightly amplified by the “happy” and “sad” illuminants.
Color does not guide eye movements: Evidence from a gaze-contingent experiment Hans Trukenbrod, Simon Barthelmé, Felix Wichmann and Ralf Engbert Color plays a crucial role in everyday life and supports actions like searching specific objects. Whether this results from modified eye guidance has not systematically been explored. Using natural scenes, we investigated the influence of color on eye movements in a visual search task. The availability of color information was limited to a constant area around fixation by presenting gaze-contingent stimuli. The remaining image was masked by a luminance-matched grayscale version of the scene. Across trials, we used six different mask sizes. Fixated stimuli ranged from black-and-white to fully colored images. Before each trial, participants were instructed to look for a bullseye-shaped target defined either by luminance or luminance plus color. Our results show that color information was not used to guide eye movements. Except for minor disruptions in conditions with small masks, fixation durations and saccade amplitudes did not differ across conditions. Even looking for a specific color target did not change statistical measures of eye guidance. While it is beyond question that color supports vision, our results suggest that color does not modify eye-movement control. While color might help to facilitate processes like object segmentation, oculomotor control seems to be unaffected by color information.
S-cone contribution to non-assimilative color spreading in the watercolor configuration Mikako Kuroki and Eiji Kimura A colored line flanking a darker contour will appear to spread its color over the area enclosed by the line (watercolor effect). The watercolor effect has been characterized as an assimilative effect, but non-assimilative spreading has also been demonstrated with the same spatial configuration; e.g., when a black inner contour is paired with a blue outer contour, yellowish spreading can be observed (Pinna & Grossberg, 2005, JOSA, 22, 2207-2221). In investigating color spreading in the watercolor configuration, we found that the two types of color spreading could be observed simultaneously but on the opposite side of the contours. The assimilative spreading occurred in the area enclosed by the lighter contour, while the non-assimilative spreading occurred in the area enclosed by the darker contour. Furthermore, we also found that S-cone signals played a critical role in the non-assimilative spreading; the induced color generally varied along the S-cone axis in the chromaticity diagram and the magnitude of the spreading correlated with differences in S-cone contrast between the inner and outer contours. These results suggest that the interaction between luminance and S-cone signals is critical for the non-assimilative spreading, and that different visual mechanisms contribute to the assimilative and non-assimilative spreading.
Variation in the vision of a rectangular sequence, interspersed in a achromatic or monochromatic grating Jorge Montalvá Colomer, Ignacio Tortajada Montañana and Mariano Aguilar Rico Like what happens with achromatic gratings and chromatic sequences (paper presented at the X National Optics Meeting in Zaragoza, Spain) the value of the Bezold effect on the vision of the sequence depends on the orientation of the grating. With red monochromatic grating, the difference between the vertical and horizontal orientations (higher value with the horizontal orientation) is almost constant on the order of 0.02 to 0.04 increasing the angular frequency of the grating. This behavior is different from the green or blue monochromatic gratings, although the value of the Bezold effect is still greater with the horizontal orientation. The behavior of this difference blue or green monochromatic gratings with the angular frequency of the grating is different from the red monochromatic grating. In blue gratings the difference increases (although slightly) with the angular frequency of the grating for frequencies below 2 cycles/degree (approximately), decreasing for higher frequencies. In green gratings the difference between the Bezold effect of two orientations (horizontal - vertical) also decreases with lower frequencies of 2 cycles/degree (approximately), greatly increasing (0.01 to 0.10) with higher frequencies.
Perceptual misbinding of colour in rivalrous motion stimuli Ryan T. Maloney, Sarah K. Lam and Colin W. G. Clifford The correct and coherent binding of distinct visual features (colour, motion, orientation, etc) is a hallmark of normal vision. The binocular rivalry paradigm is useful for studying these interactions as it provides a situation in which conflicting images at corresponding retinal locations compete for perceptual dominance. Based on the existing evidence, it has been argued that chromatic information carried by parvocellular channels tends to promote rivalry, whereas motion signals carried predominantly by magnocellular channels tend to combine interocularly [He et al., 2005, in: Binocular Rivalry, D Alais and R Blake, Cambridge, MIT Press]. Somewhat counter to this, a misbinding of isoluminant colour within rivalrous form has recently been reported, such that the dominant percept consists of colours sourced from both eyes [Hong and Shevell, 2006, Visual Neuroscience, 23, 561-566]. Investigating this effect using counter-rotating isoluminant colour gratings, we found that while stimulus motion tended to rival, colour tended to misbind: the colours of both gratings appeared to be carried in the motion of a single grating. This suggests that the determinants of rivalry are not necessarily the same for colour and motion and supports the idea that motion rivalry can occur in isolation (Alais and Parker, 2006, Neuron, 52, 911-920).
Organic Light-Emitting Diode Monitors in Vision Science Tobias Elze, Christopher Taylor and Peter Bex Organic light-emitting diode (OLED) monitors, although currently still niche products on the display market, are becoming more and more popular. While the dominant display technology, liquid crystal displays (LCDs), has been analyzed and criticized with respect to their temporal behavior, OLED devices have not yet been assessed for applicability in vision research experiments. Here, we analyze the photometric and colorimetric output of OLEDs. In contrast to LCDs, OLED monitors work without backlights and can display darker black levels. However, we found cases where the color primaries were not able to display the full luminance range of 8 bit, which may impair experiments relying on color. Moreover, we found duty cycles per frame which were substantially less than one. The resulting high amplitude frequency component at the refresh rate can be within the window of visibility of neurons in the visual system and may impair applications in visual neuroscience like single unit recordings. On the other hand, compared to LCDs, luminance transitions on OLEDs occur almost instantaneously and are independent of preceding frames. Therefore, OLEDs are considerably more suitable than LCDs for experiments that require high temporal precision. Their temporal precision is comparable to that of cathode ray tube (CRT) monitors.
Temporal discrimination of flashing colour stimuli under chromatic background Rytis Stanikunas, Algimantas Svegzda, Lina Saveikyte, Henrikas Vaitkevicius, Remigijus Bliumas, Ausra Daugirdiene and Vaiva Kulbokaite The temporal sensitivity of human visual system was investigated. The two separate lights which flash at different temporal intervals were presented under coloured background. The three primary LED’s (red, green and blue) were used as flash lights. The background extending full visual field was illuminated by one of the primary LED’s or D65 illuminant composed of three primary LED’s. Our results suggest that background colour has influence on temporal discrimination threshold increasing sensitivity for some colours and reducing it for others. Also, the background colour induces change in flash colour perception for some combinations of background and stimuli colour. It seems likely, that the temporal discrimination process has three stages. At the first stage stimuli are perceived as flashing together. At the second stage flashes are perceived as different in time, but it is not possible to tell which stimulus flashes first. At the third stage it is clearly visible which stimulus flashed first. [Supported by the Research Council of Lithuania MIP-23/2010 and MIP-013/2012]