Tilt Aftereffect Research Articles

Visual adaptation as optimal information transmission

We propose that visual adaptation in orientation, spatial frequency, and motion can be understood from the perspective of optimal information transmission. The essence of the proposal is that neural response properties at the system level should be adjusted to the changing statistics of the input so as to maximize information transmission. We show that this principle accounts for several well-documented psychophysical phenomena, including the tilt aftereffect, change in contrast sensitivity and post-adaptation changes in orientation discrimination. Adaptation can also be considered on a longer time scale, in the context of tailoring response properties to natural scene statistics. From the anisotropic distribution of power in natural scenes, the proposal also predicts differences in the contrast sensitivity function across spatial frequency and orientation, including the oblique effect.

Neural substrates of the tilt illusion.

It has been suggested that direct and indirect tilt illusions and after-effects have different mechanisms, namely that the direct effects arise in VI and are sensitive to differences in spatial and temporal parameters between test and inducing stimuli, whereas indirect effects arise in extrastriate cortex and are insensitive to such parameters. When Wolfe (Vision Research 1984; 24: 1959-64) reported that large direct tilt after-effects occurred with short test flashes, he postulated that either there are distinct mechanisms which process brief and longer duration stimuli or that there are distinct mechanisms that are not primarily concerned with duration but are differentially responsive to temporal parameters amongst several others. In three experiments we demonstrate that large direct tilt illusions can be induced when parameters other than duration are manipulated, including contrast and spatial frequency, and that such large effects can occur when stimulus parameters are chosen to favour preferentially either the transient (magnocellular-like) system or the sustained (parvocellular-like) system. These results are thus consistent with Wolfe's second hypothesis. None of these stimulus manipulations had any effect on indirect tilt illusions, consistent with previous findings and hypotheses about the different mechanisms of the direct and indirect effects.

Orientation discrimination and tilt aftereffects with luminance and illusory contours.

Orientation discrimination and tilt aftereffects (TAEs) were measured to determine if the orientation of luminance and illusory contours are processed by separate mechanisms. The assumption was made that if a single mechanism supports the perception of both types of contours, then illusory and luminance contours that support the same level of orientation discrimination will be equally effective adapting patterns. Experiment I found that orientation discrimination psychometric functions for illusory and luminance contours are similar, confirming that performance could be matched. Experiment II measured orientation discrimination for a range of intensities for both contours. Experiment III measured TAEs following adaptation to illusory and luminance contours that supported a similar range of orientation discrimination. Similar TAEs were not observed, thus rejecting the single mechanism hypothesis. Experiments IV and V sought to validate the assumption that equivalent orientation discrimination predicts equivalent TAEs by using stimuli that seemed likely to be represented by the same visual mechanism. Luminance contours masked by randomly placed dots and unmasked luminance contours were used with the same procedures as experiments II and III. Equal TAEs were not observed for masked and unmasked contours matched on orientation discrimination, suggesting the assumption relating discriminability to adaptation was incorrect.

Illusions as neuro-signs.

Visual neuroscience is generally concerned with neural and/or behavioural responses to contours. For the physical stimulus, contours are defined as differences in light intensity, or luminance, like those between black letters on white paper. One set of stimuli, however, involves seeing discontinuities of brightness where no luminance differences exist. These illusory contours can be produced by gaps in figures (like the missing sectors in the blue image above). The white triangular shapes that radiate from the centre appear brighter than the white background, even though all the white areas have the same luminance. The shapes are produced by the white sectors of the blue discs and by the terminations of the concentric circles. The illusory contours even appear to curve between the discs rather than follow the shortest straight line. They can also be induced by line terminations, as in the case of the central vase shape of the image below, where the colour of the lines appears to spread into the area so defined.Figure 3View Large Image Figure ViewerDownload Hi-res image Download (PPT) Illusory contours behave much like physical ones — they interact with one another and can produce tilt after-effects (see Curr Biol 1998, 8:R509-510). When their orientations change systematically it can be sufficient to determine a shape, as in the red image above. The illusory radiating spokes (implied by arcs of circles) are straight on one side but change direction on the other. Because the changes occur at regular points on the illusory line they can define a head in profile (the left-facing head can be seen more easily if the image is viewed out of focus); the red colour appears to extend over the whole head area. It might seem unlikely that such phenomena would direct us towards neural levels of visual processing (what I have called ‘neuro-signs’ in this series) but this could be the case. Some single cells in areas V1 and V2 of the brains of macaque monkeys respond to aligned gaps or aligned discontinuities of line endings. There are probably neuro-signs for illusory contours because many of the objects we encounter in the environment do not have boundaries which are sharply distinguishable from the background. N Wade, Department of Psychology, University of Dundee, Dundee DD1 4HN, UK.

Evidence that both area V1 and extrastriate visual cortex contribute to symmetry perception

Bilateral symmetry is common in nature and most animals seem able to perceive it. Many species use judgements of symmetry in various behaviours, including mate selection [1–3]. Originally, however, symmetry perception may have developed as a tool for generating object-centered, rather than viewer-centered, descriptions of objects, facilitating recognition irrespective of position or orientation [4]. There is evidence that the visual system treats the orientation of axes-of-symmetry in the same way it treats the orientation of luminance-defined contours [5], suggesting that axes-of-symmetry act as ‘processing tokens’ [6]. We have investigated the characteristics of neural mechanisms giving rise to the perceived orientation of axes-of-symmetry. We induced tilt aftereffects with symmetrical dot patterns, eliciting perceived angle expansion and contraction effects like those usually observed with luminance-defined contours [7,8]. Induction of aftereffects during binocular rivalry resulted in a reduction of the magnitude of these effects, consistent with the aftereffects being mediated in extrastriate visual cortex, probably between visual areas V2 and MT [9]. In a second experiment in which the aftereffects were induced monocularly, their magnitudes were measured in the unadapted eye. Contraction effects transferred completely, suggesting that they are mediated by binocular cells. Expansion effects did not transfer completely, consistent with their having a monocular component. These data suggest that information about the orientation of axes-of-symmetry may be available as early as area V1, but that processing continues in extrastriate cortex.

Edge-finding in human vision: a multi-stage model based on the perceived structure of plaids

Visual experiments on the perceived structure of plaid patterns are reported and reviewed. Plaids composed of two or three sinusoidal grating components were viewed both with and without the influence of masking or adapting patterns. The results lead to the outline of a multi-stage model of the processes by which human vision derives edge-feature descriptions from the retinal image via local spatial filters. The main proposed stages are: (i) local, oriented filtering; (ii) local combination of filter outputs across orientation (or spatial frequency) to form (iii) filtered image ‘patches’; (iv) coherent combination of patches across space to form a ‘neural image’; (v) zero-crossing analysis on the neural image to localize edge features in (vi) a feature map; (vii) application of coding rules to describe feature attributes at zero-crossing locations. Evidence for the filter combination and zero-crossing stages [(ii) and (v)] comes from experiments on the perceived structure of plaids, where the perceived location and orientation of edges was predicted by zero-crossings in the output of a circular filter, and not by the Fourier component orientations. By contrast, results on the tilt aftereffect for plaids could be understood entirely as an interaction between Fourier components, but with consequential changes in the perception of ZCs that required a consideration of ‘patchwise’ analysis and the patch combination stage [(iii) and (iv)]. Theory and evidence for coding rules (vii) that describe the blur, contrast and orientation of local edge features are discussed.

The role of implicit axes of bilateral symmetry in orientation processing

An important role for the near-effortless perception of bilateral symmetry in human (and other) observers may well be to process the orientation of symmetrical or near-symmetrical objects which do not have explicitly delineated up-down axes. From psychophysical investigations of one- and two-dimensional tilt illusions and after-effects, as well as from orientation discrimination studies, evidence is presented here which points to the existence of neural substrates which are sensitive to real (or explicit) as well as virtual (or implicit) contours.

Stages of Orientation Coding Assessed by the Tilt Aftereffect

We investigated orientation coding via the spatial-frequency tuning of the tilt aftereffect (TAE). In the single-adaptation condition, subjects adapted to single gratings of 1 or 8 cycles deg−1, 40% contrast, tilted 15° clockwise or anticlockwise from vertical; in two double-adaptation conditions the 1 and 8 cycles deg−1 gratings were superimposed at opposite orientations (‘plaid’ condition) or at the same orientation (‘parallel’ condition). Test gratings of 1, 2, 4, and 8 cycles deg−1, 20% contrast, were presented for 150 ms in an interleaved staircase procedure that measured the TAE by nulling it, hence making a tilted test grating appear vertical. Initial adaptation was for 3 min, topped up for 2 s between test presentations. Results from the single-grating condition indicated broad spatial-frequency tuning of the TAE, since the effect was still strong when tested three octaves away from the adapter. In the parallel condition, the TAEs were around the average of those reported in the single condition. Negligible TAEs were found in the 1+8 cycles deg−1 plaid condition, indicating that opposing adaptations had effectively cancelled each other out. These findings strengthen the suggestion of Olzak and Thomas (1992 Vision Research32 1885 – 1898) that orientation is encoded via an integrative mechanism which pools or sums the outputs of different spatial-frequency channels, and further imply that much of the adaptation responsible for the TAE occurs at this later broad-band stage.

A Neural Network Model of Tilt Aftereffects

RF-LISSOM, a self-organising model of laterally connected orientation maps in the primary visual cortex, was used to study the psychological phenomenon known as the tilt aftereffect. The model allows observation of activation and connection patterns between large numbers of neurons simultaneously, making it possible to relate higher-level phenomena to low-level events, which is difficult to do experimentally. In RF-LISSOM, the same self-organising processes that are responsible for the development of the orientation map and its lateral connections are shown to result in tilt aftereffects over short time scales in the adult. The results give computational support for the idea that direct tilt aftereffects arise from adaptive lateral interactions between feature detectors, as has long been surmised. They also suggest that indirect effects could result from the conservation of synaptic resources during this process. The model thus provides a unified computational explanation of self-organisation and both direct and indirect tilt aftereffects in the primary visual cortex.

From Elements to Perceptions

Gestalt psychologists in the early part of the century challenged psychophysical notions that perceptual phenomena can be understood from a punctate (‘atomistic’) analysis of the elements present in the stimulus. Their ideas also inhibited later attempts to explain vision in terms of single-unit recordings from individual neurons. A rapprochement between Gestalt phenomenology and physiology seemed unlikely when the first ECVP was held in Marburg, Germany, in 1978. Since that time, response properties of neurons have been discovered that invite an interpretation of visual phenomena (including ‘illusions’) in terms of neuronal processing. Indeed, it is now possible to understand some Gestalt phenomena on the basis of known neurophysiological mechanisms. I begin by outlining the great strides that have been made since the advent of microelectrode recording from single neurons. Initially, cells (‘detectors’) selectively responding to the contrast, spatial frequency, wavelength, orientation, movement, and disparity of a stimulus placed in their receptive fields were used to interpret simple perceptual phenomena (eg, Mach bands, Hermann grids, tilt aftereffect, MAE). In recent years, cells at higher levels of the visual system have been discovered that might explain a number of more complex phenomena: the perception of illusory (occluded) contours by end-stopped cells in area V2, the filling-in of artificial scotomata by neurons in V3, colour constancy by ‘perceptive’ neurons in V4, and the perception of coherent motion in dynamic noise patterns by cells in MT. Studies of flow fields and biological motion in area MST have recently been added to account for our perceptions as we move through our environment. Prompted by these findings, a shift from local to global interactions ‘beyond the classical receptive field’ has taken place in our search for the neural substrates of perception. Current research has focused on three kinds of mechanisms: (i) converging feed-forward projections as the basis for new response properties emerging at higher levels, (ii) recruitment of lateral connections to explain filling-in, and (iii) backward propagation from higher to lower levels to account for binding and figure - ground segregation. How such mechanisms compute large-scale surface properties such as brightness, colour, and depth from local features—indeed how they construct the surfaces themselves from complex natural scenes—is only one of the many questions that are under scrutiny today. Future research will have to tackle the all-important question: How does the analysed information come together again? Furthermore, the contributions of eye movements, attention, learning, other sense modalities, and motor actions will have to be taken into consideration before we arrive at a more complete understanding of visual perception.

Centre-Surround Tuning of Orientation Detectors in Human Vision

We present a new psychophysical method to study the tuning of orientation detectors in the human visual system. The stimulus consists of a sequence of sinusoidal gratings of random orientations and spatial phases (but of fixed spatial frequency) shown at a high presentation rate (30 Hz) in 60 s long trials. The gratings are seen through a circular aperture. The subject's task is to report, as fast as possible, by pressing a key, when the presence of a horizontal grating is seen embedded in the stimulus sequence. The data are analysed by calculating the distribution of orientations present in the stimulus sequence at different times [before] the key was pressed. Similar experiments can be done by asking the subject to detect vertical and oblique orientations. In these experiments we used 100% contrast, a 3 deg diameter circular aperture, and spatial frequencies ranging from 1 to 4 cycles deg−1. The resulting orientation-tuning profiles have a ‘centre—surround’ (or Mexican hat) shape in the orientation domain. These findings are consistent with the idea of fast ‘lateral inhibition’ between orientation detectors. The centre—surround profiles may explain systematic errors in visual angle judgments, such as the perceptual expansion of acute angles and the contraction of obtuse angles, the tilt aftereffect, and the effects observed in the Zöllner, Hering, Ponzo, and Poggendorff illusions.

Do Monocular Tilt and Spatial Frequency Aftereffects Induce the Binocular Perception of Inclination and Slant?

Prolonged viewing of a set of tilted lines can affect the perceived orientation of a second set of lines with a different physical orientation (tilt aftereffect). Similarly, prolonged viewing of a set of lines of a particular spatial periodicity can affect the perceived periodicity of a second set of lines with a different physical periodicity (spatial frequency shift aftereffect). We investigated whether a binocular difference resulting from monocular tilt or spatial frequency aftereffects could induce the perception of 3-D inclination or slant respectively. Observers adapted to monocular patterns (5 deg in diameter) arranged in a vertical ‘dumbbell’ configuration in dichoptic alternation. The adapting patterns differed in either orientation (±6.25° or ±11.25°) or in spatial frequency (±0.5 or ±0.75 octaves) from a test surface comprising vertical lines at 4 cycles deg−1. The period of adaptation was 3 min. Observers judged whether the test surfaces appeared to (i) form a convex or concave hinge in depth (after adaptation to tilt), or (ii) to slant in opposite directions about a vertical axis (after adaptation to periodicity). Using a relative slant/inclination judgment may be more sensitive than depth matching or nulling of a single surface (eg Sloane and Blake, 1987 Perception & Psychophysics42 569 – 575). Our results suggest that (i) differences in perceived periodicity in separate monocular images do not induce the impression of stereoscopic slant (confirming the results of Sloane and Blake) and (ii) differences in perceived orientation in separate monocular images do not induce an impression of stereoscopic inclination.

Do Monocular Tilt and Spatial Frequency Aftereffects Induce the Binocular Perception of Inclination and Slant?

Prolonged viewing of a set of tilted lines can affect the perceived orientation of a second set of lines with a different physical orientation (tilt AE). Similarly, prolonged viewing of a set of lines of a particular spatial periodicity can affect the perceived periodicity of a second set of lines with a different physical periodicity (spatial frequency shift AE). We investigated whether a binocular difference resulting from monocular tilt or spatial-frequency aftereffects could induce the perception of 3-D inclination or slant, respectively. Observers adapted to monocular patterns (5 deg in diameter) arranged in a vertical ‘dumbbell’ configuration in dichoptic alternation. The adapting patterns differed in either orientation (±6.25° or ±11.25°) or in spatial frequency (±0.5 or ±0.75 octaves) from a test surface comprising vertical lines at 4 cycles deg−1. The period of adaptation was 3 mins. Observers judged whether the test surface appeared to (i) form a convex or concave hinge in depth (after adaptation to tilt) or (ii) to slant in opposite directions about a vertical axis (after adaptation to periodicity). Using a relative slant/inclination judgment may be more sensitive than depth matching or nulling of a single surface (eg Sloane and Blake, 1987 Perception & Psychophysics42 569 – 575). Our results suggest that (i) differences in perceived periodicity in separate monocular images do not induce the impression of stereoscopic slant [confirming the results of Sloane and Blake (1987)] and (ii) differences in perceived orientation in separate monocular images do not induce an impression of stereoscopic inclination.

The tilt aftereffect in plaids and gratings: channel codes, local signs and “patchwise” transforms

The perceived structure of a suprathreshold plaid made from two sinusoidal gratings tilted ± 45 deg from vertical usually resembles a blurred checkerboard. Physically increasing the tilt of the components away from vertical elongates the pattern horizontally, yielding rectangular checks. We asked whether illusory tilt of the components induced by adaptation to a vertical gratin would also give apparent elongation of the checks. Using a staircase method, we found that after adaptation the component orientations had to be set 3–4 deg closer to vertical to maintain a square, checkerboard-like appearance, implying that the ± 45 deg plaid did appear elongated. This result suggests that the tilt aftereffect adaptation not only distorts the orientation of 1-D patterns but can also change the relative location of features, even when their orientation is not altered. This is not easy to explain within filter models that (often implicitly) assume each receptive field carries a fixed “local sign” indicating its position within the retinal array. We consider a representation in which localized patches of the image are encoded by the coefficients of a 2-D Fourier-like transform. With a simple subtractive model of adaptation, we show that the spatial information carried by each patch would be distorted in just the way observed for plaids and gratings. The fact that perceived structure is both distorted and coherent, not fragmented, after adaptation suggests an additional, more global process whereby local patches are combined to form a coherent, composite “neural image”. We offer a simple principle that could establish this coherence. If the positions of local patches are adjusted so as to maximize the contrast energy of the composite image, then the spatial distortions of the patches (induced by selective adaptation) are carried through to the global structure of the composite. Thus, the conflict between channel codes and local signs is resolved, but perceptual distortion is the result.

The effect of dark and equiluminant occlusion on the interocular transfer of visual aftereffects

Lehmkuhle and Fox [(1976) Vision Research, 16, 428–430] reported that interocular transfer (IOT) of a translational motion aftereffect (MAE) was greater if the non-adapting eye viewed an equiluminant field than if it viewed a dark field. They recommended equiluminant occlusion of the non-adapted eye when measuring IOT of aftereffects. We tested this proposal in three experiments. First, we assessed IOT with equiluminant and dark occlusion for three different classes of aftereffects. Although transfer was greater with equiluminant occlusion for the translational MAE, there was no significant difference in the amount of transfer for the tilt aftereffect or the contrast threshold elevation effect. Second, we tested the hypothesis that spuriously large IOT could be the result of an aftereffect from tracking eye movements in the non-adapting eye. When potential tracking movements were reduced by using rotating spokes, a rotating spiral or contracting concentric circles, there was a corresponding reduction in the occlusion-dependent transfer. Third, we found that luminance shifts had no influence on the amount of transfer when all contours were eliminated from the non-adapting eye. We conclude that the type of occlusion used for measuring IOT of the translational MAE is important only when visible contours in the non-adapting eye contribute to the adapting process.

Mechanisms of purely subjective contour tilt aftereffects

Neurones tuned for second-order stimuli—those which have edges defined by properties other than luminance and colour—have been frequently observed in prestriate cortex and in area V2 there are neurones which explicitly and unambiguously signal the orientation of purely subjective contours, i.e. contours with no Fourier components at the orientation of the perceived edge [von der Heydt, R. & Peterhans, E. (1989) Journal of Neuroscience, 9, 1731–1748]. No neurones in area V1 showed similar tuning characteristics. In addition, it has been demonstrated that like real contours, purely subjective test contours are subject to tilt aftereffects following prolonged viewing of an adapting stimulus, whether that stimulus is real or subjective [Paradiso, M. A., Shimojo, S. & Nakayama, K. (1989) Vision Research, 29, 1205–1213]. This result supports the assertion that the cortical processes responsible for real contour perception are also those giving rise to subjective contour perception. The data reported here further examined this hypothesis. Four experiments show that purely subjective contours exhibit both direct and indirect tilt aftereffects and tilt illusions like those observed with real contours. Further, they provide evidence that direct and indirect subjective contour effects, like direct and indirect real contour effects, arise via the operation of two mechanisms: a low level process, possibly lateral inhibition between orientation channels, and a second “higher-order” process. The data suggest that processing of orientation information beyond the striate cortex is similar to that which occurs in area V1 and the data are consistent with models of contour processing which assume that all perceived contours, both real and subjective, arise from a common mechanism.

Spatial vision of the achromat: The tilt aftereffect

Tilt aftereffects measured in an achromat and a trichromat control subject exhibit similar strength and angular extent, indicating normal orientation coding in the rod-based achromat vision.

Psychophysical evidence for area V2 involvement in the reduction of subjective contour tilt aftereffects by binocular rivalry.

Previous research suggests binocular rivalry disrupts extrastriate, but not striate processes, although the locus along the visual pathway at which such disruption first occurs is uncertain. It has been argued that subjective contours arise via a two-stage process in which end-stopped cells feed into orientation-sensitive neurones in V2, and that orientation aftereffects induced with subjective contours are the product of mechanisms similar to those giving rise to real contour aftereffects. If binocular rivalry disrupts the acquisition of subjective contour aftereffects, then it follows from this model that rivalry disrupts processing in V2. Experiments reported here confirm this and provide evidence which suggests binocular rivalry arises through interactions between binocular neurones, rather than via some type of specialized binocular rivalry mechanism.

On the contribution of a binocular 'AND' channel at contrast threshold.

Three experiments are reported in which an attempt was made to isolate the contribution of an AND channel by measuring aftereffects following alternating monocular adaptation. The first two were designed to test Wolf and Held's proposal that the binocular AND channel does not respond at contrast threshold. In the first experiment the relative sizes of monocular and binocular contrast threshold elevation were compared with the pattern of aftereffects obtained in a study of the suprathreshold tilt aftereffect. Identical patterns of results were obtained under the two adaptation conditions. In the second experiment, the monocular and binocular contrast-reduction aftereffect reported by Blakemore et al was measured over a wide range of reference contrasts. As in the previous experiment, the monocular effect was greater than the binocular effect. This occurred at all reference contrasts. These data support the conclusion that the AND channel contributes to visual performance in the same manner, irrespective of stimulus contrast. In the final experiment an alternative explanation for existing evidence against the existence of an AND channel was assessed.

Illusory, motion, and luminance-defined contours interact in the human visual system

Psychophysical studies of interactions between contours defined by different image attributes report that luminance-defined and illusory contours show little if any interaction. Because the contours defined by these attributes may vary in perceptual saliency, we employed the tilt aftereffect (TAE) and a cross-adaptation procedure to evaluate interaction effects between luminance-defined and illusory contours under varying saliency conditions as well as to explore the interaction between illusory and motion-defined contours. When contour salience of the adaptation or test stimuli was modified by the addition of various amounts of static noise, we observed a TAE for all combinations of contour types including the novel motion-illusory and illusory-motion pairs. The interactions demonstrated between the contour classes in this as well as other studies suggests contour invariance in the orientation domain.