Hannah E. Smithson

Sensory, computational and cognitive components of human colour constancy

When the illumination on a scene changes, so do the visual signals elicited by that scene. In spite of these changes, the objects within a scene tend to remain constant in their apparent colour. We start this review by discussing the psychophysical procedures that have been used to quantify colour constancy. The transformation imposed on the visual signals by a change in illumination dictates what the visual system must ‘undo’ to achieve constancy. The problem is mathematically underdetermined, and can be solved only by exploiting regularities of the visual world. The last decade has seen a substantial increase in our knowledge of such regularities as technical advances have made it possible to make empirical measurements of large numbers of environmental scenes and illuminants. This review provides a taxonomy of models of human colour constancy based first on the assumptions they make about how the inverse transformation might be simplified, and second, on how the parameters of the inverse transformation might be set by elements of a complex scene. Candidate algorithms for human colour constancy are represented graphically and pictorially, and the availability and utility of an accurate estimate of the illuminant is discussed. Throughout this review, we consider both the information that is, in principle, available and empirical assessments of what information the visual system actually uses. In the final section we discuss where in our visual systems these computations might be implemented.

Photostimulator allowing independent control of rods and the three cone types.

This report describes a second-generation photostimulator with four primary lights that allows independent control of the stimulation of the four receptor types in the human eye. The new design uses LEDs (with light levels controlled by eight drivers that include voltage-to-frequency converters that provide 1-micros pulses at frequencies up to 250 kHz), with four center channels being combined by use of a fiber optic assembly, and likewise for four surround channels. Four fiber optic bundles are merged into a single bundle whose output is fed into a spatial homogenizer terminated by a diffuser. An interference filter is sandwiched between each LED and the fiber optic bundle. Two camera lenses collimate light from the diffusers, one for center and one for surround. The center-surround field configuration is formed by a photometric cube with a mirrored ellipse on the hypotenuse. A field lens places images of the diffusers in the plane of an artificial pupil. The fields are highly uniform. Following alignment and calibration, the center and surround fields are indistinguishable. An observer calibration procedure, designed to compensate for prereceptoral filtering, is shown by calculation to correct also for normal observer receptoral spectral sensitivity variation. With the instrument calibrated for the individual observer, a peripherally fixated 200-ms 40% contrast rod center field pulse, highly conspicuous under dark adaptation, is invisible following light adaptation.

Colour constancy in context: roles for local adaptation and levels of reference.

By determining the locations of boundaries between colour categories, we measured changes in the colour appearance of test-reflectances as a function of the simulated illumination. Test-reflectances were displayed against a variegated background of reflectance samples. Under prolonged adaptation to each illuminant, observers demonstrated a high degree of appearance-based colour constancy. By using backgrounds that consisted of chromatically biased sets of reflectances, we tested whether this stability depends on estimates of the illuminants cone-coordinates based on simple scene statistics. The chromatic bias of the background had only a small effect on the classification of test materials. To compare the roles of spatially local and spatially extended estimation processes, we then (unknown to the observer) simulated different illuminants on the test and on the background. Observers continued to demonstrate reasonable colour constancy. To examine the relative roles of automatic adaptation and perceptual strategies, we reduced the duration of exposure to the test compared to exposure to the background (under the conflicting illuminant). The results suggest that mechanisms that preserve information across successive test-presentations (e.g. spatially local adaptation with a time course of a few seconds, and perceptual adjustments to levels of reference) are key determinants of the stability of colour appearance.

Latency characteristics of the short-wavelength-sensitive cones and their associated pathways

There are many distinct types of retinal ganglion and LGN cells that have opponent cone inputs and which may carry chromatic information. Of interest are the asymmetries in those LGN cells that carry S-cone signals: in S-ON cells, S+ signals are opposed by (L + M) whereas, in many S-OFF cells, L+ signals are opposed by (S + M), giving -S + L - M (C. Tailby, S. G. Solomon, & P. Lennie, 2008). However, the S-opponent pathway is traditionally modeled as +/-[S - (L + M)]. A phase lag of the S-cone signal has been inferred from psychophysical thresholds for discriminating combinations of simultaneous sinusoidal modulations along +/-[L - M] and +/-[S - (L + M)] directions (C. F. Stromeyer, R. T. Eskew, R. E. Kronauer, & L. Spillmann, 1991). We extend this experiment, measuring discrimination thresholds as a function of the phase delay between pairs of orthogonal component modulations. When one of the components isolates the tritan axis, there are phase delays at which discrimination is impossible; when neither component is aligned with the tritan axis, discrimination is possible at all delays. The data imply that the S-cone signal is delayed by approximately 12 ms relative to (L - M) responses. Given that post-receptoral mechanisms show diverse tuning around the tritan axis, we suggest that the delay arises before the S-opponent channels are constructed, possibly in the S-cones themselves.

Modulation of the face- and body-selective visual regions by the motion and emotion of point-light face and body stimuli.

Neural regions selective for facial or bodily form also respond to facial or bodily motion in highly form-degraded point-light displays. Yet it is unknown whether these face-selective and body-selective regions are sensitive to human motion regardless of stimulus type (faces and bodies) or to the specific motion-related cues characteristic of their proprietary stimulus categories. Using fMRI, we show that facial and bodily motions activate selectively those populations of neurons that code for the static structure of faces and bodies. Bodily (vs. facial) motion activated body-selective EBA bilaterally and right but not left FBA, irrespective of whether observers judged the emotion or color-change in point-light angry, happy and neutral stimuli. Facial (vs. bodily) motion activated face-selective right and left FFA, but only during emotion judgments for right FFA. Moreover, the strength of responses to point-light bodies vs. faces positively correlated with voxelwise selectivity for static bodies but not faces, whereas the strength of responses to point-light faces positively correlated with voxelwise selectivity for static faces but not bodies. Emotional content carried by point-light form-from-motion cues was sufficient to enhance the activity of several regions, including bilateral EBA and right FFA and FBA. However, although the strength of emotional modulation in right and left EBA by point-light body movements was related to the degree of voxelwise selectivity to static bodies but not static faces, there was no evidence that emotional modulation in fusiform cortex occurred in a similarly stimulus category-selective manner. This latter finding strongly constrains the claim that emotionally expressive movements modulate precisely those neuronal populations that code for the viewed stimulus category.

Do masks terminate the icon

Iconic memory is operationally defined by part-report experiments (Sperling, 1960). If a mask is presented after the target, the mask is thought to be superposed on the target in the iconic representation, or to displace it from the representation. But could a cue presented after a pattern mask still allow selection within the target array? A target array of letters was followed by a checkerboard mask. We compared two target–mask interstimulus intervals (ISIs; 0 and 100 ms), and six cue delays. At ISI = 0 ms, performance was at chance, for part report and whole report. At ISI = 100 ms, with the shortest cue delay, observers demonstrated a part-report advantage of 25–30%. As cue delay increased the part-report advantage decreased. These results are inconsistent with an iconic memory that is automatically displaced or overwritten by new information. We consider two alternatives: a second-stage store, which represents letters in terms of their high-level features and which the mask cannot penetrate, or a four-dimensional store that preserves separately the representations of the target and its aftercoming mask. We discuss the implications of our results for studies that use backward masking to “terminate the icon”.

The loss of the PDE6 deactivating enzyme, RGS9, results in precocious light adaptation at low light levels

The GTPase activating protein, RGS9-1, is vital for the deactivation and regulation of the phototransduction cascade (C. K. Chen et al., 2000; C. W. Cowan, R. N. Fariss, I. Sokal, K. Palczewski, & T. G. Wensel, 1998; W. He, C. W. Cowan, & T. G. Wensel, 1998; A. L. Lyubarsky et al., 2001). Its loss through genetic defects in humans has been linked to a slow recovery to changes in illumination (K. M. Nishiguchi et al., 2004). Such a deficit is to be expected because RGS9-1 normally speeds up the deactivation of the activated phosphodiesterase effector molecule, PDE6*, and thus accelerates the turning off of the visual response. Paradoxically, however, we find that the cone response in an observer lacking RGS9-1 is faster at lower light levels than it is in a normal observer. Though surprising, this result is nonetheless consistent with molecular models of light adaptation (e.g., E. N. Pugh, S. Nikonov, & T. D. Lamb, 1999), which predict that the excess of PDE6* resulting from the loss of RGS9-1 will shorten the visual integration time and speed up the visual response at inappropriately low light levels. The gain in speed caused by the superfluity of PDE6* at lower light levels compensates for the loss caused by its slow deactivation; thus quickening the response relative to that in the normal. As the light level is increased and the PDE6* concentration in the normal rises relative to that in the observer lacking RGS9-1, the temporal advantage of the latter is soon lost, leaving only the deficit due to delayed deactivation.

Compatible and incompatible representations in visual sensory storage

Sensory storage shows a short-lived part-report advantage that survives an aftercoming visual noise pattern (Smithson & Mollon, 2006). We tested whether such an advantage survives different types of high-contrast mask. The target was a 3 × 4 array of digits. The mask could be (a) a noise pattern, (b) an array of eights, or (c) an array of random digits. In a preliminary experiment, target and mask were interleaved (at 140 Hz) and target contrast was varied to determine the level at which performance fell to chance. In the main experiment, target and mask were separated by an inter-stimulus-interval (ISI) of 100, 150, or 200 ms. An auditory part-report cue that was presented 240 ms after target offset supported a part-report advantage at all ISIs for noise masks, at ISIs greater than 100 ms for digit-8 masks, but not at any ISI for random-number masks. Increasing cue delay, in the range 240 to 730 ms, produced a decline in the advantages we measured. The differences in part-report superiority with different types of mask call for revision of the model of visual sensory storage as a single canvas on which successive items are superposed. When mask and target are sufficiently different, a representation of low-contrast target digits can be maintained independently of the representation of an aftercoming, high-contrast mask. However, when the same target is followed by a mask composed of high-contrast random digits, an independent representation of the target does not remain available for access.

Vision science and adaptive optics, the state of the field.

Adaptive optics is a relatively new field, yet it is spreading rapidly and allows new questions to be asked about how the visual system is organized. The editors of this feature issue have posed a series of question to scientists involved in using adaptive optics in vision science. The questions are focused on three main areas. In the first we investigate the use of adaptive optics for psychophysical measurements of visual system function and for improving the optics of the eye. In the second, we look at the applications and impact of adaptive optics on retinal imaging and its promise for basic and applied research. In the third, we explore how adaptive optics is being used to improve our understanding of the neurophysiology of the visual system.

A three-dimensional color space from the 13th century.

We present a new commentary on Robert Grossetestes De colore, a short treatise that dates from the early 13th century, in which Grosseteste constructs a linguistic combinatorial account of color. In contrast to other commentaries (e.g., Kuehni & Schwarz, Color Ordered: A Survey of Color Order Systems from Antiquity to the Present, 2007, p. 36), we argue that the color space described by Grosseteste is explicitly three-dimensional. We seek the appropriate translation of Grossetestes key terms, making reference both to Grossetestes other works and the broader intellectual context of the 13th century, and to modern color spaces.