Search Results

The purpose of this chapter is twofold. First, after a brief consideration of the major computational problems that must be resolved by the visual stages of reading, it reviews the evidence for or against the hypothesis that part of the left inferotemporal cortex functions as a specialized “visual word form area” in literate adults. Second, assuming that this area does indeed contain neurons that have become attuned to visual word recognition, it asks how these neurons might be organized. Considering that word recognition probably arises from a minimal “recycling” of the preexisting visual object recognition system, the chapter presents a hypothetical model, according to which a hierarchy of local combination detector (LCD) neurons, inspired from the known architecture of the primate inferotemporal cortex, collectively implement an invariant neural code for written words.

As with most complex behaviors, visual word recognition is thought to result from the dynamic interplay between the elements of a distributed cortical and subcortical network. To understand fully how visual word recognition is achieved therefore, and how it may fail in developmental dyslexia, not only the necessary and sufficient complement of nodes that comprise this network—its functional anatomy—need to be identified, but also how information flows through this network with time needs to be understood, and indeed how the structure of the network itself may adapt in both the short and long term. This chapter takes a historical approach to reviewing recent magnetoencephalography (MEG) research that elucidates these temporal dynamics, focusing particularly on events with the first 300 milliseconds (ms) of a visually presented word, and which should set crucial constraints on models of visual word recognition and reading.

This chapter reviews the principles of assessment of perceptual disorders, auditory recognition deficits, tactile recognition deficits, and visual object recognition deficits. The implicit assumption is that visual object recognition skills have a hierarchical and parallel organization. In ordinary clinical practice however, object recognition disorders are ‘random’. The question arises as to where a clinician should start when assessing a patient with object and/or face recognition problems. The first consideration is to obtain information on potential sensory problems, visuospatial neglect, aphasia, etc. that may at least partly account for perceptual/agnosic deficits. As a general rule, clinicians should be alert to the complaints and errors patients make during other tasks involving picture interpretation. If the patient fails these tests, there is no need for tests of visual agnosias either of the aperceptive or associative type — any agnosia diagnosed in this context would be ‘pseudoagnosia’.

This chapter addresses the varying definitions of “modularity” assumed by different fields (e.g., philosophy, psychology, neuroscience), and focuses on evaluating Fodor's notion of information encapsulation. It is shown that attentional instructions can modulate low-level visual processes, and that visual input of a moving face can modulate the auditory perception of a phoneme. In fact, cortical regions in the ferret's brain that normally receive auditory input can learn to accommodate incoming synapses from the optic tract. Thus, although it is clear that various anatomical regions of the brain are somewhat specialized for specific perceptual abilities, the fluidity and ubiquity with which they interact in real-time indicates that cognitive processes, such as spatial attention, visual event recognition, and speech perception, exhibit not modularity but instead something that might be called distribularity.

The brain processes involved in visual object recognition are described. Evidence is presented that what is computed are sparse distributed representations of objects that are invariant with respect to transforms including position, size, and even view in the ventral stream towards the inferior temporal visual cortex. Then biologically plausible unsupervised learning mechanisms that can perform this computation are described that use a synaptic modification rule what utilises a memory trace. These are compared with deep learning and other machine learning approaches that require supervision.

This chapter aims to formulate a neurodynamical theory and model that addresses the issues of how spatial and object attention mechanisms can be integrated and can function as a unitary system in visual search and visual recognition tasks. An important novel idea in this model is that the dorsal stream and the ventral stream interact at multiple points and levels, and the locus of intersection is a function of the scale of analysis. The model can reproduce the findings of a number of attention-related neurophysiological experiments and it provides a united conceptual framework to account for several apparently disparate psychological processes such as spatial and object attention, object recognition and localization, and serial and parallel search.

This chapter shows that the feature hierarchy approach has a number of advantages in performing object recognition over other approaches (see Section 25.3), and that some of the key computational issues that arise in these architectures have solutions (see Sections 25.4 and 25.50. The neurophysiological and computational approach taken here focuses on a feature hierarchy model in which invariant representations can be built by self-organizing learning based on the statistics of the visual input. The model can use temporal continuity in an associative synaptic learning rule with a short-term memory trace, and/or it can use spatial continuity in Continuous Spatial Transformation learning. The model of visual processing in the ventral cortical stream can build representations of objects that are invariant with respect to translation, view, size, and lighting. The model uses a feature combination neuron approach with the relative spatial positions of the objects specified in the feature combination neurons, and this provides a solution to the binding problem. The model has been extended to provide an account of invariant representations in the dorsal visual system of the global motion produced by objects such as looming, rotation, and object-based movement. The model has been extended to incorporate top-down feedback connections to model the control of attention by biased competition in for example spatial and object search tasks (see further Chapter 6). The model has also been extended to account for how the visual system can select single objects in complex visual scenes, how multiple objects can be represented in a scene, and how invariant representations of single objects can be learned even when multiple objects are present in the scene. The model has also been extended to account for how the visual system can select multiple objects in complex visual scenes using a simulation of saliency computations in the dorsal visual system, and then with fixations on the salient parts of the scene perform view-invariant visual object recognition using the simulation of the ventral visual stream, VisNet. It has also been suggested in a unifying proposal that adding a fifth layer to the model and training the system in spatial environments will enable hippocampus-like spatial view neurons or place cells to develop, depending on the size of the field of view (Section 24.3.11). We have thus seen how many of the major computational issues that arise when formulating a theory of object recognition in the ventral visual system (such as feature binding, invariance learning, the recognition of objects when they are in cluttered natural scenes, the representation of multiple objects in a scene, and learning invariant representations of single objects when there are multiple objects in the scene), could be solved in the cortex, with tests of the hypotheses performed by simulations that are consistent with complementary neurophysiological results. The approach described in this chapter is unifying in a number of ways. First, a set of simple organizational principles involving a hierarchy of cortical areas with convergence from stage to stage, and competitive learning using a modified associative learning rule with a short-term memory trace of preceding neuronal activity, provide a basis for understanding much processing in the ventral visual stream, from V1 to the inferior temporal visual cortex. Second, the same principles help to understand some of the processing in the dorsal visual stream by which invariant representations of the global motion of objects may be formed. Third, the same principles continued from the ventral visual stream onwards to the hippocampus help to show how spatial view and place representations may be built from the visual input. Fourth, in all these cases, the learning is possible because the system is able to extract invariant representations because it can utilize the spatio-temporal continuities and statistics in the world that help to define objects, moving objects, and spatial scenes. Fifth, a great simplification and economy in terms of brain design is that the computational principles need not be different in each of the cortical areas in these hierarchical systems, for some of the important properties of the processing in these systems to be performed. The principles of cortical operation that are illustrated include the following: One is that advantage is taken of the statistics of inputs from the world to help learning, with for example temporal and spatial continuity being relevant. Another is that neurons need to learn to respond to non-linear combinations of their inputs, in the case of vision including their spatial arrangement which is provided by the convergent topology from area to area of the visual cortex, using principles such as competitive learning. Another principle is that the system must be able to form sparse distributed representations with neurons that encode perceptual and invariance properties, so that the next stage of cortical processing can read the information using dot product decoding as in a pattern associator, autoassociator, or competitive network. Another principle is the use of hierarchical cortical computation with convergence from stage to stage, which breaks the computation down into neuronally manageable computations. Another principle is breaking the computation down into manageable parts, by for example not trying to analyze the whole of a scene simultaneously, but instead using successive fixations to objects in different parts of the scene, and maintaining in short-term memory a limited representation of the whole scene. In conclusion, we have seen in this chapter how a major form of perception, the invariant recognition of objects, involves not only the storage and retrieval of information, but also major computations to produce invariant representations. Once these invariant representations have been formed, they are used for many processes including not only recognition memory (see Section 24.2.6), but also associative learning of the rewarding and punishing properties of objects for emotion and motivation (see Chapter 24), the memory for the spatial locations of objects and rewards (see Chapter 24), the building of spatial representations based on visual input (Section 24.3.11), and as an input to short-term memory (Section 4.3.1), attention (Chapter 6), and decision systems (Section 5.6).

This chapter presents the theory of visual attention (TVA). It begins with a review of choice models for visual recognition (categorization) and for visual search (partial report). Roughly speaking, these models are non-process models; they provide descriptive equations with strong empirical constraints, but make little or no attempt to specify the temporal course of the information processing underlying performance. It then considers a race model for selection from multi-element displays. The race model is a process model; it specifies temporal characteristics of processing. Further, the race model is perfectly consistent with the descriptive-choice model for visual search, which can simply be derived from it mathematically. Finally, the unified theory of visual recognition and attentional selection (TVA) is developed by integrating choice models for recognition into the race model framework. In a mathematical sense, TVA includes the previous models as special cases, and hence inherits their success in accounting for empirical findings. TVA is not only a process model, but is also computational; it specifies the computations by which selection is supposed to be done.

This chapter suggests that the demands on processing that are placed by the particular stimuli that are processed and by the particular tasks that are performed play crucial roles in determining which subsystem recognizes objects and supports repetition priming effects. The recognition and priming of incomplete objects in particular should be informative for testing this possibility. Thus, the chapter summarizes evidence from studies of priming for incomplete objects that support the hypothesized influence of stimulus and task demands.

One of the most remarkable findings to arise from investigations of visual recognition impairments due to brain injury has been that patients who do not seem to show normal, overt recognition of things they see may none the less demonstrate a form of non-conscious, covert recognition if appropriate tests are used. These patients do not seem to suffer any general alteration of consciousness, but one specific aspect, awareness of recognition, is lost. This chapter provides a tutorial review of findings of covert recognition, concentrating on those arising in cases of prosopagnosia, but sketching their relation to other neuropsychological phenomena. It then considers some of the issues which arise and accounts of the phenomena that have been attempted, in the context of recent findings.

This chapter explores the hypothesis that the brain routinely makes affective predictions during visual recognition. It suggests that the brain's prediction about the meaning of visual sensations of the present includes some representation of the affective impact of those (or similar) sensations in the past. An affective prediction, in effect, allows the brain to anticipate and prepare to act on those sensations in the future. Furthermore, it is hypothesized that affective predictions are made quickly and efficiently, only milliseconds after visual sensations register on the retina. From this perspective, sensations from the body are a dimension of knowledge—they help people to identify what an object is when upon encountering it, based in part on past reactions. If this hypothesis is correct, then affective responses signaling an object's salience, relevance, or value, do not occur as a separate step after the object is identified—affective response assists in seeing an object as what it is from the very moment that visual stimulation begins.

Modern studies indicate several different causes of visual object agnosia. Lissauer himself realized this at the end of the nineteenth century; he proposed that visual recognition can be separated into apperceptive and associative stages, and that each when impaired had its own characteristic agnosia. The apperceptive stage would correspond to the final stage of purely perceptual processing; it was considered to be intact if a brain-injured person could accurately copy items they could not recognize. The associative stage would give the percept meaning by linking it to previous experience; leading to the widely adopted description of associative agnosia as corresponding to a normal percept ‘stripped of its meaning’.

This chapter examines some neural processes that a scene image undergoes as it moves through the visual system. It focuses on two opposite yet highly interactive neural systems, the frontoparietal network and the ventral visual stream. Visual recognition mechanisms in the ventral stream lean toward certain objects in visual scenes because they occupy a space that has already been allotted for a high priority by the lateral intraparietal area and the frontal eye fields. While the ventral visual system processes and determines the objects in that environment, the frontoparietal network allocates and points visual attention to important features of the environment.This division of labor by the two systems is supported by the view that spatial selection and target identification are separable parts of finding objects in visual scenes.

An intriguing set of issues arises from work in cognitive neuropsychiatry, and these are touched on in this chapter. They include what makes people's percepts normally seem so real to them. Delusions in which the feeling of things being real seems to be dysfunctional also raise important questions about the sense of self and self-existence. This chapter examines how visual recognition mechanisms support the experience of reality by looking at some of the consequences of recognition impairments, and especially those caused by brain injury. This involves considering how one's emotional reactions to visual stimuli relate to their ability to recognize them, and examining preserved non-conscious aspects of recognition in cases of face recognition impairment after brain injury (prosopagnosia), everyday recognition errors, and delusional misidentifications in psychiatric and neuropsychiatric patients.

This chapter begins with the historical development of comparative neuropsychology. George Ettlinger was one of the few who actively combined human and animal research, and he did so consistently throughout his scientific career. He thus embodied the biological study of brain and behaviour with a commitment matched by few others. In this volume, an attempt has been made to assemble authoritative essays by current experts in a number of the areas that interested George Ettlinger during his career, though because of the sheer breadth of those interests, this aim has not been wholly achieved. His earliest concerns, relating to the nature of visual agnosia and the brain systems underlying visual recognition, are addressed in subsequent chapters.

This introductory chapter first sets out the book's purpose, which is to examine the representations and stereotypes of South Asian Americans in relation to immigrant narratives of assimilation in American film and television. It theorizes the performance of accent as a means of representing race and particularly national origin beyond visual identification. For South Asians, accent simultaneously connotes difference and privilege. To focus on an Indian vocal accent is to reconsider racialization predicated on visual recognition. The remainder of the chapter discusses vocal accents and racial hierarchies; South Asian American and Indian American identities; popular Culture, Orientalism, and racial performance; and comedy and racial performance. It concludes with an overview of the subsequent chapters.

Cognitive brain disorders, such as dyslexia, also come in for their fair share of misconceptions, and this is the case among laypeople, policymakers, and even teachers. But oddly, the misconceptions surrounding disorders affecting “cold” cognition are the mirror images of our errors in reasoning about psychiatric disorders that we associate with “warm” emotions. We are all too eager to assume that major depression is innate, controlled by the material brain and out of the patients’ hands; for dyslexia, we incorrectly assume the opposite—that it is “just” “in our mind” (not in our brain and genes). Chapter 13 explains how reading works and what dyslexia really is, and it shows how our misconceptions about it arise from Dualism and Essentialism.