A motor theory of speech perception, initially proposed to account for results of early experiments with synthetic speech, is now extensively revised to accommodate recent findings, and to relate the assumptions of the theory to those that might be made about other perceptual modes. According to the revised theory, phonetic information is perceived in a biologically distinct system, a ‘module ’ specialized to detect the intended gestures of the speaker that are the basis for phonetic categories. Built into the structure of this module is the unique but lawful relationship between the gestures and the acoustic patterns in which they are variously overlapped. In consequence, the module causes perception of phonetic structure without translation from preliminary auditory impressions. Thus, it is comparable to such other modules as the one that enables an animal to localize sound. Peculiar to the phonetic module are the relation between perception and production it incorporates and the fact that it must compete with other modules for the same stimulus variations.

this paper we consider basic representations based on localized functions that provide an overcomplete representation of the space of interest. In most cases the amplitudes can be thought of as samples of the PDF and the expansion functions as interpolating formulas, such as the generalized radial basis functions used by Poggio and Girosi [3] ( see also the discussion in Hertz et al, section 6.4 [2]).

Through vision, we derive a rich understanding... This article reviews some computational studies of vision, focusing on edge detection, binocular stereo, motion analysis, intermediate vision and object recognition.

Neurons in the visual cortex of monkeys respond selectively to the disparity between the images in the two eyes. Recent recordings have shown that some of the disparity-selective neurons in the primary visual cortex and the posterior parietal cortex are modulated by the distance of fixation. A population of such gain-modulated, disparity-selective neurons forms a set of basis functions of horizontal disparity and distance of fixation that can be used as an intermediate representation for computing egocentric distance. This distributed representation is consistent with psychophysical studies of human depth perception; in contrast, neurons explicitly tuned to distance are not consistent with how we perceive distance. In a population model that includes noise in the firing rates of neurons, the perceived distance is

between images presented to the two eyes induce a strong sensation of depth. More recent experiments with random-dot stereograms have shown that disparity is a sufficient cue for stereopsis (Julesz 1960, 1971). Disparity-tuned neurons in visual cortex were first demonstrated in the cat (Barlow et al. 1967; Nikara

Ever since the time of Wheatstone (1838) it has been known that disparities between images presented to the two eyes will induce a strong sensation of depth More recently, the studies of Julesz (1960, 1971) have shown

Visual experience is intrinsically subjective. The manifest unity of our perceptions belies the indeterminacy of our sensations. A single pattern of excitation on the retinal receptors is consistent with many possible worlds of objects. The simple problem of determining object brightness exemplifies the ambiguities inherent in the retinal image. The photons incident on a given receptor may indicate the presence of a bright or dark object depending on the overall level of illumination. In mediating automatic gain control, the horizontal cells of the retina express an assumption about the nature of the illuminant. These neurons in the most peripheral part of the visual system take the first step toward interpretation of the image. By a process of lateral inhibition, they discount the effect of illumination and determine the brightness of an object relative to that of nearby objects. At all levels of complexity, the visual system interprets each data point within the context of the scene. This interpretation is consistent with the input data and with the internal structure of the system. Through evolution, the structure of the visual system provides correspondence

Previous models of stereopsis have concentrated on the task of binocularly matching left and right eye primith'es uniquely. A disparity smoothness constraint is often invoked to limit the number of possible matches. These approaches neglect the fact that surface discontinuities are both abundant in natural eve!)'day scenes, and provide a useful cue for scene segmentation. da Vinci stereopsis refers to the more general problern of dealing Ivith surface discontinuities and their associated unmatched monocular regions}vithin binocular scenes. This study develops a mathematical realization of a neural netlvork theory of biological vision, called FACADE theo!)', that sholvs holY early cortical stereopsis processes are related to later cortical processes of three-dimensional surface representation. The mathematical model demonstrates through computer simulation holY the visual cortex Inay generate threedilnensional boundary segmentations and use them to control filling-in of three-dimensional surface properties in response to visual scenes. Model mechanisms correctly match disparate binocular regions while filling-in monocular regions Ivith the correct depth within a binocularly viewed scene. This achievelnent required the introduction of a new multiscale binocular filter for;\'tereo matching}vhich clarifies holv cortical complex cells match image contours of like contrast polarity, while pooling signals from opposite contrast polarities. The filter also suggests how false binocular matches and unmatched monocular cues are automatically handled across multiple spatial scales. Pooling of signals froln even- and odd-S)'lnmetric simple cells at complex cells helps to eliminate spurious activity peaks in matchable