Perception

With each eye fixation, we experience a richly detailed visual world. Yet recent work on visual integration and change direction reveals that we are surprisingly unaware of the details of our environment from one view to the next: we often do not detect large changes to objects and scenes (‘change blindness’). Furthermore, without attention, we may not even perceive objects (‘inattentional blindness’). Taken together, these findings suggest that we perceive and remember only those objects and details that receive focused attention. In this paper, we briefly review and discuss evidence for these cognitive forms of ‘blindness’. We then present a new study that builds on classic studies of divided visual attention to examine inattentional blindness for complex objects and events in dynamic scenes. Our results suggest that the likelihood of noticing an unexpected object depends on the similarity of that object to other objects in the display and on how difficult the priming monitoring task is. Interestingly, spatial proximity of the critical unattended object to attended locations does not appear to affect detection, suggesting that observers attend to objects and events, not spatial positions. We discuss the implications of these results for visual representations and awareness of our visual environment.

Research has suggested that novice drivers have different search strategies compared with their more experienced counterparts, and that this may contribute to their increased accident liability. One issue of concern is whether experienced drivers have a wider field of peripheral vision than less experienced drivers. This study attempted to distinguish between people of varying driving experience on the basis of their functional fields of view. Participants searched video clips taken from a moving driver's perspective for potential hazards while responding to peripheral target lights. Hit rates for peripheral targets decreased for all participant groups as processing demands increased (ie when hazards occurred) and as the eccentricity of the target increased, though there was no interaction. An effect of experience was also found which suggests that this paradigm measures a perceptual skill or strategy that develops with driving experience.

Given a constant stream of perceptual stimuli, how can the underlying invariances associated with a given input be learned? One approach consists of using generic truths about the spatiotemporal structure of the physical world as constraints on the types of quantities learned. The learning methodology employed here embodies one such truth: that perceptually salient properties (such as stereo disparity) tend to vary smoothly over time. Unfortunately, the units of an artificial neural network tend to encode superficial image properties, such as individual grey-level pixel values, which vary rapidly over time. However, if the states of units are constrained to vary slowly, then the network is forced to learn a smoothly varying function of the training data. We implemented this temporal-smoothness constraint in a backpropagation network which learned stereo disparity from random-dot stereograms. Temporal smoothness was formalised with the use of regularisation theory by modifying the standard cost function minimised during training of a network. Temporal smoothness was found to be similar to other techniques for improving generalisation, such as early stopping and weight decay. However, in contrast to these, the theoretical underpinnings of temporal smoothing are intimately related to fundamental characteristics of the physical world. Results are discussed in terms of regularisation theory and the physically realistic assumptions upon which temporal smoothing is based.

Humans understand mechanical events to involve physical bodies interacting by contact, but intentional events involve agents that can also interact at a distance. We investigated infant sensitivity to causality in a simple event in which one agent appears to react to another without contact. Infants 9 months old were habituated to one of two events involving a computer-animated red square moving nonrigidly—like a caterpillar—towards a green square. In the ‘reaction event’, the green object moved in turn before the red one stopped, while in the ‘pause event’ the green object moved after the red one stopped. After habituation, each infant saw the habituation movie played in reverse. This test involved identical spatiotemporal changes for reaction and pause event, but the reversed reaction additionally involved a change in the causal roles. Infants dis-habituated to reversal of the reaction but not the pause event, a result which suggests sensitivity to causation-at-a-distance. This ability could support development of social cognition and theory of mind.

Trompe l'oeil pictures have been produced for hundreds of years. They attempt to create the impression of a surface that has different three-dimensional structure to the work; successful examples of trompe l'oeil typically constrain the observer's viewpoint and require use of a single eye. The works of Patrick Hughes are in relief but are painted to appear like conventional flat pictures; those parts that protrude from the picture plane are pictorially distant, or in reverse perspective. Movements of the observer result in fluid distortions of the pictorial image. These distortions occur with binocular observation and over a wide range of viewing distances.

When a small frontoparallel surface (a test strip) is surrounded by a larger slanted surface (an inducer), the test strip is perceived as slanted in the direction opposite to the inducer. This has been called the depth-contrast effect, but we call it the slant-contrast effect. In nearly all demonstrations of this effect, the inducer's slant is specified by stereoscopic signals; and other signals, such as the texture gradient, specify that it is frontoparallel. We present a theory of slant estimation that determines surface slant via linear combination of various slant estimators; the weight of each estimator is proportional to its reliability. The theory explains slant contrast because the absolute slant of the inducer and the relative slant between test strip and inducer are both estimated with greater reliability than the absolute slant of the test strip. The theory predicts that slant contrast will be eliminated if the signals specifying the inducer's slant are consistent with one another. It also predicts reversed slant contrast if the inducer's slant is specified by nonstereoscopic signals rather than by stereo signals. These predictions were tested and confirmed in three experiments. The first showed that slant contrast is greatly reduced when the stereo-specified and nonstereo-specified slants of the inducer are made consistent with one another. The second showed that slant contrast is eliminated altogether when the stimulus consists of real planes rather than images on a display screen. The third showed that slant contrast is reversed when the nonstereo-specified slant of the inducer varies and the stereo-specified slant is zero. We conclude that slant contrast is a byproduct of the visual system's reconciliation of conflicting information while it attempts to determine surface slant.

Petter's rule applies to two-dimensional patterns formed by two overlapping surfaces that alternately appear in front of one another. It states that the surface with the shorter contours in the region where the surfaces look superimposed has a greater probability of appearing in front of the other surface. An experiment is reported the results of which show that Petter's rule is valid for chromatically homogeneous and for uniformly dense dotted patterns, and invalid for different kinds of chromatically inhomogeneous patterns. Petter's rule has been found to be valid when the overlapping surfaces have contours with gaps. It is proposed that Petter's rule derives from the dynamics of filling-in of contour gaps.

The structure of human disparity representation is examined through (i) adaptation experiments and (ii) model simulations of the data. Section 3 presents results of adaptation experiments designed to illuminate the structure of human disparity representation. Section 4 presents model simulations of three different disparity representation schemes. In the experiments, participants adapted to a 0.133 cycle deg⁻¹ sinusoidally corrugated surface with 10 min of arc peak-to-trough disparity. A flat test surface was briefly presented, in which the aftereffect surface was perceived. Adapt and test surfaces were placed on disparity pedestals and thus presented in front of or behind the plane of fixation. The adapt surface could be offset from the fixation plane by ±8 to 24 min of arc. The test surface could be offset from the fixation plane by ±8 to 48 min of arc. The depth aftereffect was measured in different disparity planes by a nulling method and ‘topping-up’ procedure. Aftereffect tuning functions were obtained whose bandwidths, magnitudes, and tuning depended on the disparity planes of both the adapt and test surfaces. These parameters were used to constrain the models tested in section 4. On the basis of the two studies, it is argued that the human stereoscopic system encodes spatial changes of disparity using channels localised within disparity planes. A localised disparity-gradient model of the human representation of disparity is proposed.

The processing of facial line drawings was investigated in either simultaneous or sequential matching trials with either the same or different viewpoint, showing pictures of faces either in the same modes (both photographs or line drawings) or different modes (one in each mode). Line drawings were particularly difficult to match in memory rather than under perceptual conditions, and line drawings did not allow the creation of efficient structural codes. These deficits of line representations underline the assumption that the face-processing system is inflexible when it is confronted with edge-based material.