Alexa I. Ruppertsberg

Rendering complex scenes for psychophysics using RADIANCE: How accurate can you get?

Rendering packages are used by visual psychophysicists to produce complex stimuli for their experiments, tacitly assuming that the simulation results accurately reflect the light-surface interactions of a real scene. RADIANCE is a physically based, freely available, and commonly used rendering software. We validated the calculation accuracy of this package by comparing simulation results with measurements from real scenes. RADIANCE recovers color gradients well but the results are shifted in color space. Currently, there is no better simulation alternative for achieving physical accuracy than by combining a spectral rendering method with RADIANCE.

The chromatic input to global motion perception.

For over 30 years there has been a controversy over whether color-defined motion can be perceived by the human visual system. Some results suggest that there is no chromatic motion mechanism at all, whereas others do find evidence for a purely chromatic motion mechanism. Here we examine the chromatic input to global motion processing for a range of color directions in the photopic luminance range. We measure contrast thresholds for global motion identification and simple detection using sparse random-dot kinematograms. The results show a discrepancy between the two chromatic axes: whereas it is possible for observers to perform the global motion task for stimuli modulated along the red-green axis, we could not assess the contrast threshold required for stimuli modulated along the yellowish-violet axis. The contrast required for detection for both axes, however, are well below the contrasts required for global motion identification. We conclude that there is a significant red-green input to global motion processing providing further evidence for the involvement of the parvocellular pathway. The lack of S-cone input to global motion processing suggests that the koniocellular pathway mediates the detection but not the processing of complex motion for our parameter range.

When S-cones contribute to chromatic global motion processing

There is common consensus now that color-defined motion can be perceived by the human visual system. For global motion integration tasks based on isoluminant random dot kinematograms conflicting evidence exists, whether observers can (Ruppertsberg et al., 2003) or cannot (Bilodeau & Faubert, 1999) extract a common motion direction for stimuli modulated along the isoluminant red-green axis. Here we report conditions, in which S-cones contribute to chromatic global motion processing. When the display included extra-foveal regions, the individual elements were large ( approximately 0.3 degrees ) and the displacement was large ( approximately 1 degrees ), stimuli modulated along the yellowish-violet axis proved to be effective in a global motion task. The color contrast thresholds for detection for both color axes were well below the contrasts required for global motion integration, and therefore the discrimination-to-detection ratio was >1. We conclude that there is significant S-cone input to chromatic global motion processing and the extraction of global motion is not mediated by the same mechanism as simple detection. Whether the koniocellular or the magnocellular pathway is involved in transmitting S-cone signals is a topic of current debate (Chatterjee & Callaway, 2002).

Creating physically accurate visual stimuli for free: Spectral rendering with RADIANCE

Visual psychophysicists, who study object, color, and light perception, have a demand for software that produces complex but, at the same time, physically accurate stimuli for their experiments. The number of computer graphic packages that simulate the physical interaction of light and surfaces is limited, and mostly they require the purchase of a license. RADIANCE (Ward, 1994), however, is freely available and popular in the visual perception community, making it a prime candidate. We have shown previously that RADIANCE’S simulation accuracy is greatly improved when color is coded by spectra, rather than by the originally envisaged RGB triplets (Ruppertsberg & Bloj, 2006). Here, we present a method for spectral rendering with RADIANCE to generate hyperspectral images that can be converted to XYZ images (CIE 1931 system) and then to machine-dependent RGB images. Generating XYZ stimuli has the added advantage of making stimulus images independent of display devices and, thereby, facilitating the process of reproducing results across different labs. Materials associated with this article may be downloaded from www.psychonomic.org.

Sensitivity to luminance and chromaticity gradients in a complex scene

Image gradients--smooth changes in color and luminance--may be caused by intrinsic surface reflectance properties or extrinsic illumination phenomena, including shading, shadowing, and inter-reflections. In turn, image gradients may provide the visual system with information concerning the origin of these factors, such as the orientation of surfaces with respect to the light source. Color gradients induced by mutual illumination (MI) may play a similar role to that of luminance gradients in shape-from-shading algorithms; it has been shown that 3D shape perception modulates the influence of MI on surface color perception (M. G. Bloj, D. Kersten, & A. C. Hurlbert, 1999). In this study, we assess human sensitivity to changes in color and luminance gradients that arise from changes in the light source position, within a complex scene. In Experiment 1, we tested whether observers were able to discriminate between gradients due to different light source positions. We found that observers reliably detected a change in the gradient information when the light source position differed by only 4 deg from the reference scene. This sensitivity was mainly based on the luminance information in the gradient (Experiments 2 and 3). Some observers make use of the spatial distribution of chromaticity and luminance values within gradients when discriminating between them (Experiment 4). The high sensitivity to gradient differences supports the notion that gradients contain information that may assist in the recovery of 3D shape and scene configuration properties.

Reflecting on a room of one reflectance.

We present a numerical analysis of rendered pairs of rooms, in which the spectral power distribution of the illuminant in one room matched the surface reflectance function in the other room, and vice versa. We ask whether distinction between the rooms is possible and on what cues this discrimination is based. Using accurately rendered three-dimensional (3D) scenes, we found that room pairs can be distinguished based on indirect illumination, as suggested by A. L. Gilchrist and A. Jacobsen (1984). In a simulated color constancy scenario, we show that indirect illumination plays a pivotal role as areas of indirect illumination undergo a smaller appearance change than areas of direct illumination. Our study confirms that indirect illumination can play a critical role in surface color recovery and shows how computer rendering programs, which model the light-object interaction according to the laws of physics, are valuable tools that can be used to analyze and explore what image information is available to the visual system from 3D scenes.

The integration of local chromatic motion signals is sensitive to contrast polarity.

Global motion integration mechanisms can utilize signals defined by purely chromatic information. Is global motion integration sensitive to the polarity of such color signals? To answer this question, we employed isoluminant random dot kinematograms (RDKs) that contain a single chromatic contrast polarity or two different polarities. Single-polarity RDKs consisted of local motion signals with either a positive or a negative S or L-M component, while in the different-polarity RDKs, half the dots had a positive S or L-M component, and the other half had a negative S or L-M component. In all RDKs, the polarity and the motion direction of the local signals were uncorrelated. Observers discriminated between 50% coherent motion and random motion, and contrast thresholds were obtained for 81% correct responses. Contrast thresholds were obtained for three different dot densities (50, 100, and 200 dots). We report two main findings: (1) dependence on dot density is similar for both contrast polarities (+S vs. -S, +LM vs. -LM) but slightly steeper for S in comparison to LM and (2) thresholds for different-polarity RDKs are significantly higher than for single-polarity RDKs, which is inconsistent with a polarity-blind integration mechanism. We conclude that early motion integration mechanisms are sensitive to the polarity of the local motion signals and do not automatically integrate information across different polarities.

Visual sensitivity to achromatic gradients with different luminance profiles

A calibrated monitor controlled by a 42-bit graphics card (CRS VISAGE) was used to display the stimuli. Detection experiment (3 observers) •There were 6 conditions: 3 types of stimuli with different luminance profiles (Figure 2) were used with 2 different types of background (uniform or non-uniform; Figure 1). •At each presentation of the stimuli, the little squares of the non-uniform background changed randomly.

From scene to screen: colour accuracy in computer simulations used for psychophysics

In order to produce a complex three-dimensional stimulus for psychophysical research, investigators make use of computer graphics software. If we want to extrapolate the conclusions of these studies to human behaviour in the ‘real world’ we must also show that behaviour is equivalent in the simulated and the real environment. A first step towards this goal is to validate the physical accuracy of the simulation program with respect to a real scene by measuring the different physical characteristics of the scene and comparing them directly with the values produced by the simulation for an equivalent physical and geometrical set up. RADIANCE [Ward 1994], a physically-based, freely available rendering package, has been used in the computer graphics community for some time and is rapidly becoming the preferred rendering tool of the visual perception community to simulate experimental scenes for psychophysical and behavioural studies. Studies that set out to evaluate RADIANCE place emphasis on luminance accuracy but not on colour accuracy and the evidence on luminance accuracy is divided. In our opinion, RADIANCE still awaits a full spectral validation.

Global motion processing: the Red-Green mechanism

The interaction of colour and motion cues for global motion integration across space has only recently been studied. To establish the tolerance limits of global motion we first determine global motion detection thresholds (81%) as a function of the chromatic contrast in the isoluminant cone-opponent colour-space (S-(M -L) space). We further testwhether global motion in the isoluminant plane is mediated by more than one chromatic mechanism.SM-LLumWP0°90°180°270°-90°+90°