Visual illusions in neural networks: Line neutralization,tilt after effect,and angle expansion

Certain visual illusions occur in neural networks that are capable of storing partially contrasted enhanced spatial patterns in short term memory (STM), and whose feature detectors are interconnected by nontrivial generalization gradients. These include neutralization, or adaptation, of nearly vertical or horizontal lines, tilt after-effect of successively viewed lines, and perceived angle expansion. Neutralization can be achieved by networks whose vertical and horizontal representations have higher saturation levels, broader tuning curves, or stronger input pathways. Tilt after-effect and angle expansion can be achieved by shunting lateral inhibition that causes an outward peak shift in the orientationally-coded STM pattern. The amount of outward peak shift is also dependent on the size of the potassium equilibrium point. Differences between the directions of tilt aftereffect (successive contrast) and angle expansion (simultaneous contrast) are ascribed to a normalization of total activity in the STM buffer whereby present stimuli and representations in STM of past stimuli interact to form a consistent action-oriented consensus.     

Turning the other cheek: the viewpoint dependence of facial expression after-effects

How do we visually encode facial expressions? Is this done by viewpoint-dependent mechanisms representing facial expressions as two-dimensional templates or do we build more complex viewpoint independent three-dimensional representations? Recent facial adaptation techniques offer a powerful way to address these questions. Prolonged viewing of a stimulus (adaptation) changes the perception of subsequently viewed stimuli (an after-effect). Adaptation to a particular attribute is believed to target those neural mechanisms encoding that attribute. We gathered images of facial expressions taken simultaneously from five different viewpoints evenly spread from the three-quarter leftward to the three-quarter rightward facing view. We measured the strength of expression after-effects as a function of the difference between adaptation and test viewpoints. Our data show that, although there is a decrease in after-effect over test viewpoint, there remains a substantial after-effect when adapt and test are at differing three-quarter views. We take these results to indicate that neural systems encoding facial expressions contain a mixture of viewpoint-dependent and viewpoint-independent elements. This accords with evidence from single cell recording studies in macaque and is consonant with a view in which viewpoint-independent expression encoding arises from a combination of view-dependent expression-sensitive responses.     

Auditory motion aftereffects of low- and high-frequency sound stimuli

Adaptation of listeners to approaching or receding sound stimuli continued for 5 s under free-field conditions. Motion of the adaptive and test sound stimuli was simulated by means of oppositely directed linear changes in the amplitude of the low- and high-frequency noises (0.05–1 and 3–20 kHz, respectively) from two stationary loudspeakers. In a group of eight subjects with normal hearing, the auditory motion after-effect of the approaching and receding sound stimuli was evaluated by integrated indices that characterized the shift of the psychometric curves in response to the test stimuli under various conditions of listening. The aftereffect occurs in the case when the spectral composition of the adaptive and test stimuli coincides. In response to the high-frequency stimuli, the effect of adaptation to both the approaching and receding sound stimuli was observed, while in response to the low-frequency stimuli, only the approach of stimuli caused an aftereffect. There was no radial motion aftereffect in the case of mismatching the spectral bands of the adaptive and test stimuli. Thus, the frequency selectivity was characteristic of the auditory aftereffect of adaptation to the approaching and receding sound stimuli.     

Space and time in visual context

No sensory stimulus is an island unto itself; rather, it can only properly be interpreted in light of the stimuli that surround it in space and time. This can result in entertaining illusions and puzzling results in psychological and neurophysiological experiments. We concentrate on perhaps the best studied test case, namely orientation or tilt, which gives rise to the notorious tilt illusion and the adaptation tilt after-effect. We review the empirical literature and discuss the computational and statistical ideas that are battling to explain these conundrums, and thereby gain favour as more general accounts of cortical processing.     

Erratum

AN acute angle appears to be less acute than it really is^1,2. The effect has an obvious similarity to the tilt after-effect, the only difference in procedure being that the lines forming the acute angle are presented simultaneously in one case and successively in the other. Nevertheless, Blakemore et al.² consider that the effect they studied is not the tilt after-effect, on the grounds of different temporal properties: their effect builds up and dissipates very rapidly, which, they argue, is inconsistent with known adaptation phenomena, such as Gibson after-effects^3,4, which have long time constants.     

Slow and fast visual motion channels have independent binocular-rivalry stages

We have previously reported a transparent motion after-effect indicating that the human visual system comprises separate slow and fast motion channels. Here, we report that the presentation of a fast motion in one eye and a slow motion in the other eye does not result in binocular rivalry but in a clear percept of transparent motion. We call this new visual phenomenon 'dichoptic motion transparency' (DMT). So far only the DMT phenomenon and the two motion after-effects (the 'classical' motion after-effect, seen after motion adaptation on a static test pattern, and the dynamic motion after-effect, seen on a dynamic-noise test pattern) appear to isolate the channels completely. The speed ranges of the slow and fast channels overlap strongly and are observer dependent. A model is presented that links after-effect durations of an observer to the probability of rivalry or DMT as a function of dichoptic velocity combinations. Model results support the assumption of two highly independent channels showing only within-channel rivalry, and no rivalry or after-effect interactions between the channels. The finding of two independent motion vision channels, each with a separate rivalry stage and a private line to conscious perception, might be helpful in visualizing or analysing pathways to consciousness.     

Modelling fast forms of visual neural plasticity using a modified second-order motion energy model

The Adelson-Bergen motion energy sensor is well established as the leading model of low-level visual motion sensing in human vision. However, the standard model cannot predict adaptation effects in motion perception. A previous paper Pavan et al.(Journal of Vision 10:1–17, 2013) presented an extension to the model which uses a first-order RC gain-control circuit (leaky integrator) to implement adaptation effects which can span many seconds, and showed that the extended model’s output is consistent with psychophysical data on the classic motion after-effect. Recent psychophysical research has reported adaptation over much shorter time periods, spanning just a few hundred milliseconds. The present paper further extends the sensor model to implement rapid adaptation, by adding a second-order RC circuit which causes the sensor to require a finite amount of time to react to a sudden change in stimulation. The output of the new sensor accounts accurately for psychophysical data on rapid forms of facilitation (rapid visual motion priming, rVMP) and suppression (rapid motion after-effect, rMAE). Changes in natural scene content occur over multiple time scales, and multi-stage leaky integrators of the kind proposed here offer a computational scheme for modelling adaptation over multiple time scales.     

Electrophysiological investigation of the properties of the electroreceptors (ampullae of Lorenzini) in Black Sea rays

Responses of electroreceptors (ampullae of Lorenzini) in Black Sea rays to electrical stimuli were recorded in vivo as spike activity of single nerve fibers. Depending on their functional properties the fibers could be divided into silent, those with regular activity (10–15 spikes/sec) and those with grouped activity. Electrical stimuli evoked a tonic response with a varied degree of adaptation in the nerve fibers. The threshold currents were between 10^?10 and 10^?11 A/mm². The minimal latent period of the on-responses to pulses of current of maximal intensity was 15–40 msec, whereas that of the off-responses was 15–200 msec. The effect of intensity, duration, and polarity of the stimuli on the responses of the receptors and the adaptation of the electroreceptors during application of a steady current were investigated. The properties of the ampullae of Lorenzini were compared with those of other types of electroreceptors.     

ADAPTATION OF CUTANEOUS TACTILE RECEPTORS : VI. INHIBITORY EFFECTS OF POTASSIUM AND CALCIUM

1. Both solutions of Ringer plus fifteen times the normal K content, and solutions of Ringer plus fifteen times the normal Ca content markedly hasten the adaptation of single freely branching axon endings in frog''s skin to repetitive air puff stimuli. 2. The K effect is produced more rapidly than is that of Ca. The K effect is reversible by washing with Ringer''s solution, while the Ca effect is not. The Ca inhibition can, however, be reversed and recovery effected by washing with K rich solutions. 3. Evidence is discussed which indicates that Ca probably plays no rôle in normal adaptation, and the experiments are interpreted as substantiating the hypothesis of adaptation due to K.     

Fractal rotation isolates mechanisms for form-dependent motion in human vision

Here, we describe a motion stimulus in which the quality of rotation is fractal. This makes its motion unavailable to the translation-based motion analysis known to underlie much of our motion perception. In contrast, normal rotation can be extracted through the aggregation of the outputs of translational mechanisms. Neural adaptation of these translation-based motion mechanisms is thought to drive the motion after-effect, a phenomenon in which prolonged viewing of motion in one direction leads to a percept of motion in the opposite direction. We measured the motion after-effects induced in static and moving stimuli by fractal rotation. The after-effects found were an order of magnitude smaller than those elicited by normal rotation. Our findings suggest that the analysis of fractal rotation involves different neural processes than those for standard translational motion. Given that the percept of motion elicited by fractal rotation is a clear example of motion derived from form analysis, we propose that the extraction of fractal rotation may reflect the operation of a general mechanism for inferring motion from changes in form.     

HAMP domain structural determinants for signalling and sensory adaptation in Tsr,the Escherichia coli serine chemoreceptor

HAMP domains mediate input–output transactions in many bacterial signalling proteins. To clarify the mechanistic logic of HAMP signalling, we constructed Tsr‐HAMP deletion derivatives and characterized their steady‐state signal outputs and sensory adaptation properties with flagellar rotation and receptor methylation assays. Tsr molecules lacking the entire HAMP domain or just the HAMP‐AS2 helix generated clockwise output signals, confirming that kinase activation is the default output state of the chemoreceptor signalling domain and that attractant stimuli shift HAMP to an overriding kinase‐off signalling state to elicit counter‐clockwise flagellar responses. Receptors with deletions of the AS1 helices, which free the AS2 helices from bundle‐packing constraints, exhibited kinase‐off signalling behaviour that depended on three C‐terminal hydrophobic residues of AS2. We conclude that AS2/AS2′ packing interactions alone can play an important role in controlling output kinase activity. Neither kinase‐on nor kinase‐off HAMP deletion outputs responded to sensory adaptation control, implying that out‐of‐range conformations or bundle‐packing stabilities of their methylation helices prevent substrate recognition by the adaptation enzymes. These observations support the previously proposed biphasic, dynamic‐bundle mechanism of HAMP signalling and additionally show that the structural interplay of helix‐packing interactions between HAMP and the adjoining methylation helices is critical for sensory adaptation control of receptor output.     

Perceived direction of motion determined by adaptation to static binocular images

In Li and Atick's [1, 2] theory of efficient stereo coding, the two eyes' signals are transformed into uncorrelated binocular summation and difference signals, and gain control is applied to the summation and differencing channels to optimize their sensitivities. In natural vision, the optimal channel sensitivities vary from moment to moment, depending on the strengths of the summation and difference signals; these channels should therefore be separately adaptable, whereby a channel's sensitivity is reduced following overexposure to adaptation stimuli that selectively stimulate that channel. This predicts a remarkable effect of binocular adaptation on perceived direction of a dichoptic motion stimulus [3]. For this stimulus, the summation and difference signals move in opposite directions, so perceived motion direction (upward or downward) should depend on which of the two binocular channels is most strongly adapted, even if the adaptation stimuli are completely static. We confirmed this prediction: a single static dichoptic adaptation stimulus presented for less than 1 s can control perceived direction of a subsequently presented dichoptic motion stimulus. This is not predicted by any current model of motion perception and suggests that the visual cortex quickly adapts to the prevailing binocular image statistics to maximize information-coding efficiency.     

A functional angle on some after-effects in cortical vision

The question of how our brains and those of other animals code sensory information is of fundamental importance to neuroscience research. Visual illusions offer valuable insight into the mechanisms of perceptual coding. One such illusion, the tilt after-effect (TAE), has been studied extensively since the 1930s, yet a full explanation of the effect has remained elusive. Here, we put forward an explanation of the TAE in terms of a functional role for adaptation in the visual cortex. The proposed model accounts not only for the phenomenology of the TAE, but also for spatial interactions in perceived tilt and the effects of adaptation on the perception of direction of motion and colour. We discuss the implications of the model for understanding the effects of adaptation and surround stimulation on the response properties of cortical neurons.     

Stimulus change detection in phasic auditory units in the frog midbrain: frequency and ear specific adaptation

Neural adaptation, a reduction in the response to a maintained stimulus, is an important mechanism for detecting stimulus change. Contributing to change detection is the fact that adaptation is often stimulus specific: adaptation to a particular stimulus reduces excitability to a specific subset of stimuli, while the ability to respond to other stimuli is unaffected. Phasic cells (e.g., cells responding to stimulus onset) are good candidates for detecting the most rapid changes in natural auditory scenes, as they exhibit fast and complete adaptation to an initial stimulus presentation. We made recordings of single phasic auditory units in the frog midbrain to determine if adaptation was specific to stimulus frequency and ear of input. In response to an instantaneous frequency step in a tone, 28 % of phasic cells exhibited frequency specific adaptation based on a relative frequency change (delta-f = ±16 %). Frequency specific adaptation was not limited to frequency steps, however, as adaptation was also overcome during continuous frequency modulated stimuli and in response to spectral transients interrupting tones. The results suggest that adaptation is separated for peripheral (e.g., frequency) channels. This was tested directly using dichotic stimuli. In 45 % of binaural phasic units, adaptation was ear specific: adaptation to stimulation of one ear did not affect responses to stimulation of the other ear. Thus, adaptation exhibited specificity for stimulus frequency and lateralization at the level of the midbrain. This mechanism could be employed to detect rapid stimulus change within and between sound sources in complex acoustic environments.     

The segregation and integration of colour in motion processing revealed by motion after-effects

Analysis of the colour and motion of objects is widely believed to take place within segregated processing pathways in the primate visual system. However, it is apparent that this segregation cannot remain absolute and that there must be some capacity for integration across these sub-modalities. In this study, we have assessed the extent to which colour constitutes a separable entity in human motion processing by measuring the chromatic selectivity of two kinds of after-effect resulting from motion adaptation. First, the traditional motion after-effect, where prolonged inspection of a unidirectional moving stimulus results in illusory motion in the opposite direction, was found to exhibit a high degree of chromatic selectivity. The second type of after-effect, in which motion adaptation induces misperceptions in the spatial position of stationary objects, was completely insensitive to chromatic composition. This dissociation between the chromatic selectivities of these after-effects shows that chromatic inputs remain segregated at early stages of motion analysis, while at higher levels of cortical processing there is integration across chromatic, as well as achromatic inputs, to produce a unified perceptual output.     

Visual associative memory and the orientation-contingent color after-effect

The traditional explanation of the McCollough effect (ME) by selective adaptation of single detectors selective to color and orientation suffers from a number of inconsistencies: 1) the ME lasts much longer (from several days up to 3 months) than the ordinary adaptation, the decay of the effect being completely arrested by night sleep or occluding the eye for a long time; 2) the strength of the ME practically does not depend on the intensity of adapting light; and 3) a set of related pattern-contingent after-effects discovered later required for such an explanation new detectors, specific for other patterns. These properties can be explained, however, in the framework of associative memory and novelty filters. A computational model has been developed, which consists of 1) an input layer of two (left and right eyes) square matrices with two analog receptors (red and green) in each pixel, 2) an isomorphic associative neural layer, each analog neuron being synaptically connected with all receptors of both eyes, and 3) an output layer (novelty filter). The modification of synaptic efficacies conforms to the Hebb learning rule. The function of the model was examined by simulation. After a few presentations of colored gratings, the model displays the ME that is slowly destroyed by subsequent presentations of random pictures. With a sufficiently large receptor matrix, the effect lasts a thousand times longer than the period of adaptation. Continuous darkness does not change the strength of the effect. Like in real ME, the model does not display interocular transfer. The model can account for different pattern-contingent color after-effects without assuming any predetermined specific detectors. Such detectors are constructed in the course of adaptation to specific stimuli (gratings).     

Bayesian calibration of simultaneity in audiovisual temporal order judgments

After repeated exposures to two successive audiovisual stimuli presented in one frequent order, participants eventually perceive a pair separated by some lag time in the same order as occurring simultaneously (lag adaptation). In contrast, we previously found that perceptual changes occurred in the opposite direction in response to tactile stimuli, conforming to bayesian integration theory (bayesian calibration). We further showed, in theory, that the effect of bayesian calibration cannot be observed when the lag adaptation was fully operational. This led to the hypothesis that bayesian calibration affects judgments regarding the order of audiovisual stimuli, but that this effect is concealed behind the lag adaptation mechanism. In the present study, we showed that lag adaptation is pitch-insensitive using two sounds at 1046 and 1480 Hz. This enabled us to cancel lag adaptation by associating one pitch with sound-first stimuli and the other with light-first stimuli. When we presented each type of stimulus (high- or low-tone) in a different block, the point of simultaneity shifted to "sound-first" for the pitch associated with sound-first stimuli, and to "light-first" for the pitch associated with light-first stimuli. These results are consistent with lag adaptation. In contrast, when we delivered each type of stimulus in a randomized order, the point of simultaneity shifted to "light-first" for the pitch associated with sound-first stimuli, and to "sound-first" for the pitch associated with light-first stimuli. The results clearly show that bayesian calibration is pitch-specific and is at work behind pitch-insensitive lag adaptation during temporal order judgment of audiovisual stimuli.     

Sensory adaptation

Adaptation occurs in a variety of forms in all sensory systems, motivating the question: what is its purpose? A productive approach has been to hypothesize that adaptation helps neural systems to efficiently encode stimuli whose statistics vary in time. To encode efficiently, a neural system must change its coding strategy, or computation, as the distribution of stimuli changes. Information theoretic methods allow this efficient coding hypothesis to be tested quantitatively. Empirically, adaptive processes occur over a wide range of timescales. On short timescales, underlying mechanisms include the contribution of intrinsic nonlinearities. Over longer timescales, adaptation is often power-law-like, implying the coexistence of multiple timescales in a single adaptive process. Models demonstrate that this can result from mechanisms within a single neuron.     

Auditory aftereffects of approaching and withdrawing sound sources: Dependence on the trajectory and location of adapting stimuli

Perception of approaching and withdrawing sound sources and their action on auditory aftereffects were studied in the free field. Motion of adapting stimuli was mimicked in two ways: (1) simultaneous opposite changes of amplitude of broadband noise impulses at two loudspeakers placed at 1.1 and 4.5 m from the listener; (2) an increase or a decrease of amplitude of broadband noise impulses in only one loudspeaker, the nearer or the remote one. Motion of test stimuli was mimicked in the former way. Listeners determined direction of the test stimuli motion without any adaptation (control) or after adaptation to stationary, slowly moving (with an amplitude change of 2 dB) and rapidly moving (amplitude change of 12 dB) stimuli. Percentages of “withdrawal” reports were used for construction of psychometric curves. Three phenomena of auditory perception were observed. In the absence of adaptation, a growing-louder effect was revealed, i.e., listeners reported more frequently the test sounds as the approaching ones. Once adapted to stationary or slowly moving stimuli, listeners showed a location-dependent aftereffect. Test stimuli were reported as withdrawing more often as compared with control. The effect was associated with the previous one and was weaker when the distance to the loudspeaker producing adapting stimuli was greater. After adaptation to rapidly moving stimuli, a motion aftereffect was revealed. In this case, listeners reported a direction of test stimuli motion as being opposite to that of adapting stimuli. The motion aftereffect was more pronounced when the adapting stimuli motion was mimicked in the former way, as this method allows estimation of their trajectory. There was no relationship between the motion aftereffect and the growing-louder effect, whichever way the adapting stimuli were produced. There was observed a tendency for reduction of aftereffects of approaching and for intensification of aftereffects of withdrawal with growing distance from source of adapting stimuli.     

Increased visual after-effects in migraine following pattern adaptation extend to simultaneous tilt illusion

Much previous research into visual processing in migraine supports a model of abnormal cortical processing, in between the headache attacks, that is characterised by hyperexcitability, heightened responsiveness, a lack of habituation and/or a lack of intra-cortical inhibition. Shepherd (2001) reported two adaptation studies that challenged this view, one using the tilt after-effect, the second using the motion after-effect. Models of cortical function in migraine based on hyperexcitability and a lack of inhibition lead to specific predictions in an adaptation study: there should have been smaller after-effects in people with migraine than in people without. Both experiments, however, revealed larger after-effects in the migraine group than in the control group. Here, these results are extended to the simultaneous tilt illusion and an identical pattern of results was obtained: there were consistently larger effects in the migraine group than in the control group. The results from the three experiments are not consistent with a lack of inhibition in migraine. The results are discussed in terms of alternative models of cortical function, including a lack of excitation and reduced central energy reserves.