We propose a computational theory of consciousness and model data from three experiments in visual perception. The central idea of our theory is that the contents of consciousness correspond to temporally stable states in an interconnected network of specialized computational modules. Each module incorporates a relaxation search that is concerned with achieving semantically well-formed states. We claim that being an attractor of the relaxation search is a necessary condition for awareness. We show that the model provides sensible explanations for the results of three experiments, and makes testable predictions. The first experiment (Marcel, 1980) found that masked, ambiguous prime words facilitate lexical decision for targets related to either prime meaning, whereas consciously perceived primes facilitate only the meaning that is consistent with prior context. The second experiment (Fehrer and Raab, 1962) found that subjects can make detection responses in constant time to simple visual stimuli regardless of whether they are consciously perceived or masked by metacontrast and not consciously perceived. The third experiment (Levy and Pashler, 1996) found that visual word recognition accuracy is lower than baseline when an earlier speeded response was incorrect, and higher than baseline when the early response was correct, consistent with a causal relationship between conscious perception and subsequent processing.

The ventral visual pathway implements object recognition and categorization in a hierarchy of processing areas with neuronal selectivities of increasing complexity. The presence of massive feedback connections within this hierarchy raises the possibility that normal visual processing relies on the use of computational loops. It is not known, however, whether object recognition can be performed at all without such loops (i.e., in a purely feed-forward mode). By analyzing the time course of reaction times in a masked natural scene categorization paradigm, we show that the human visual system can generate selective motor responses based on a single feed-forward pass. We confirm these results using a more constrained letter discrimination task, in which the rapid succession of a target and mask is actually perceived as a distractor. We show that a masked stimulus presented for only 26 msec---and often not consciously perceived---can fully determine the earliest selective motor responses: The neural representations of the stimulus and mask are thus kept separated during a short period corresponding to the feedforward "sweep." Therefore, feedback loops do not appear to be "mandatory" for visual processing. Rather, we found that such loops allow the masked stimulus to reverberate in the visual system and affect behavior for nearly 150 msec after the feed-forward sweep. &

TABLE OF CONTENTS 2.1. Visual areas are defined by receptive field tuning properties 2.2. Combining the distributed information

: The recently introduced Static and Dynamic State (SDS) Feedback control scheme together with its modified form, the Data Compression and Reconstruction (DCR) architecture that performs pseudoinverse computation, suggests a unified model of cortical processing including consciousness. The constraints of the model are outlined here and the features of the cortical architecture that are suggested and sometimes dictated by these constraints are listed. Constraints are imposed on cortical layers, e.g., (1) the model prescribes a connectivity substructure that is shown to fit the main properties of the `basic neural circuit' of the cerebral cortex (Shepherd and Koch [1], Douglas and Martin [2] In: The synaptic organization of the brain, Oxford University Press, 1990), and (2) the stability requirements of the pseudoinverse method offer an explanation for the columnar organization of the cortex. Constraints are also imposed on the hierarchy of cortical areas, e.g., the proposed control arc...

Introduction Contrary to popular belief, the ability of people to detect single changes in a scene can be very poor, even when all the information is visibly available to them (Grimes, 1996). Some types of changes are more easily detected than others; however, what constitutes a change sufficiently significant to be detected is unclear (O'Regan, 1992). Changes usually made to images are the removal of an object, the change in location or structure of an object, or the change of the color of an object. In a recent experiment, Rensink et al. (1997) showed that subjects are faster at determining the exact change if it is a color change than if it is a disappearance change, and in turn subjects are faster at determining an object disappearance change than an object location change. The poor ability of people to detect changes in visual scenes has been shown in experiments in which images are altered during saccades (Grimes, 1996), and in experiments in which the original image and its alte

We sought to determine whether or not motion-from-texture mechanisms have access to monocular input. Adopting a strategy used by Kolb and Braun (1995), we created drifting textures that were invisible to purely binocular processes. Monocular signals readily conveyed motions defined by local orientation and flicker. However, when left-eye and right-eye signals were displayed simultaneously, only flicker motion was visible. We conclude that motion-from-texture mechanisms do not have access to monocular input. Further evidence suggests that motion from texture involves attentional tracking. Motion from texture Attentional tracking 2nd-order motion

When the right eye’s target is the left eye’s distracter and vice versa, orientation-defined search is impossible unless, as we show here, the elements are close together. More than 1 s was required to find inverse-cyclopean texture boundaries when elements were arranged on a 16 x 16 grid. Less than 250 ms was required for a 24 x 24 grid covering the same area. The conventional view is that binocular rivalry requires at least 200 ms to develop, but our results suggest a more rapid access to monocular signals. We call this rapid form of access “proto-rivalry.”

These three experiments examined noticing rates of an unexpected object (UO) that appeared during a dynamic aircraft threat evaluation task that required participants to shift their visual attention between multiple task-relevant locations. Experiment 1 manipulated the location at which the UO appeared; no effects on noticing rates were found. However, eye-tracking data revealed trends for UOs to be noticed more when participants were looking at locations closer to where the UO appeared, or when they were making more eye-movements while the UO was present. Eye-tracking data also showed a strong link between making an eye movement to the UO and noticing it. Experiment 2 manipulated the color, direction and speed of the UO to make it more or less similar to task-relevant objects. Also, to-be-ignored (TBI) aircraft were either present or absent for each participant. An interaction between the color of the UO and the presence of TBI aircraft was found with noticing rates being greater for uniquely-colored UO’s only when no TBI aircraft were present. No overall effect of UO and target aircraft similarity was found. Experiment 3 manipulated the visual complexity and cognitive difficulty of the task. Noticing rates were higher only in the visually-simple, cognitively-3