young animal or child perceives and identifies features in its envi-, roument in an apparently effortless way. No presently known algorithms even approach this flexible, generalpurpose perceptual capability. Discovering the principles that may underlie perceptual processing is important both for neuroscience and for the development of synthetic perceptual systems. Two important aspects of the mystery of perception are (1) What processing functions does the neural “machinery ” perform on perceptual input, and what is the circuitry that implements these functions? (2) How does this “machinery ” come to be? Unlike conventional computer hardware, neural circuitry is not hard-wired or specified as an explicit set of point-to-point connections. Instead it develops under the influence of a genetic specification and. epigenetic factors, such as electrical activity, both before and after birth. How this happens is in large part unknown. Biological development processes are far too complex to hope that a relatively complete understanding of how a perceptual system develops and functions will soon emerge. But we are familiar with complex synthetic systems, such as computers, whose principles of organization can be understood without one’s knowing How can a perceptual system develop to recognize specific features of its environment, without being told which features it should analyze, or even whether its identifications are correct? in detail how the components work. Furthermore, the same principles can be used to build computers in any of several different technologies. Might there be organizing principles (1) that explain some essential aspects of how a perceptual system develops and functions; (2) that we can attempt to infer without waiting for far more detailed experimental information; and (3) that can lead to profitable experimental programs, testable predictions, and applications to synthetic perception as well as neuroscientific understanding? I believe the answer is yes, and that the use of theoretical neural networks that embody biologically-motivated rules and constraints is a powerful tool in this study. This optimism is encouraged by recent work ’ in which I have found that a multilayered network, developing according to simple yet biologically plausible “Hebbtype” rules, * self-organizes to produce

This paper extends to two dimensions the frame criterion developed by Daubechies for one-dimensional wavelets, and it computes the frame bounds for the particular case of 2D Gabor wavelets. Completeness criteria for 2D Gabor image representations are important because of their increasing role in many computer vision applications and also in modeling biological vision, since recent neurophysiological evidence from the visual cortex of mammalian brains suggests that the filter response profiles of the main class of linearly-responding cortical neurons (called simple cells) are best modeled as a family of self-similar 2D Gabor wavelets. We therefore derive the conditions under which a set of continuous 2D Gabor wavelets will provide a complete representation of any image, and we also find self-similar wavelet parameterizations which allow stable reconstruction by summation as though the wavelets formed an orthonormal basis. Approximating a "tight frame" generates redundancy which allows low-resolution neural responses to represent high-resolution images, as we illustrate by image reconstructions with severely quantized 2D Gabor coefficients. Index Terms---Gabor wavelets, coarse coding, image representation, visual cortex, image reconstruction.

We describe a novel approach to curve inference based on curvature information. The inference procedure is divided into two stages: a trace inference stage, to which this paper is devoted, and a curve synthesis stage, which will be treated in a separate paper. It is shown that recovery of the trace of a curve requires estimating local models for the curve at the same time, and that tangent and curvature information are sufficient. These make it possible to specify powerful constraints between estimated tangents to a curve, in terms of a neigh-borhood relationship called cocircularity and between curvature esti-mates, in terms of a curvature consistency relation. Because all curve information is quantized, special care must be taken to obtain accurate estimates of trace points, tangents and curvatures. This issue is ad-dressed specifically by the introduction of a smoothness constraint and a maximum curvature constraint. The procedure is applied to two types of images, artificial images designed to evaluate curvature and noise sensitivity, and natural images.

This article indicates how competition between afferent data and learned feedback expectancies can stabilize a developing code by buffering committed populations of detectors against continual erosion by new environmental demands. Tille gating phenomena that result lead to dynamically maintained critical peri(Jlds, and to attentional phenomena such as overshadowing in the adult. The fuillctional unit of cognitive coding is suggested to be an adaptive resonance, or amplification and,prolongation of neural activity, that occurs when afferent data and efferent expectancies reach consensus through a matching process. The resonant state embodies the perceptual event, or attentional focus, and its amplified and sustained activities are capable of driving slow changes of long-term memor:r"' Mismatch between afferent data and efferent expectancies yields a global sulppression of activity and triggers a reset of short-term memory, as well as raJ~id parallel search and hypothesis testing for uncommitted cells. These mechanisms help to explain and predict, as manifestations of the unified theme of stable code development, positive and negative aftereffects, the McCollough effect, spatial frequency adaptation, monocular rivalry, binocular rivalry and hysteresis, pattern completion, and Gestalt switching; analgesia, partial reinforcement acquisition effect, conditioned reinforcers, underaroused versus overaroused depression; the contingent negative variation, P300, and pontoge]lliculo-occipital waves; olfactory coding, corticogeniculate feedback, matching of proprioceptive and terminal motor maps, and cerebral dominance. The psychophysiological mechanisms that unify these effects are inherently nonlinear and parallel and are inequivalent to the computer, probabilistic, and linear models currently in use.

this paper, we propose a Bayesian theory of hierarchical cortical computation based both on (a) the mathematical and computational ideas of computer vision and pattern the- ory and on (b) recent neurophysiological experimental evidence. We ,2 have proposed that Grenander's pattern theory 3 could potentially model the brain as a generafive model in such a way that feedback serves to disambiguate and 'explain away' the earlier representa- tion. The Helmholtz machine 4, 5 was an excellent step towards approximating this proposal, with feedback implementing priors. Its development, however, was rather limited, dealing only with binary images. Moreover, its feedback mechanisms were engaged only during the learning of the feedforward connections but not during perceptual inference, though the Gibbs sampling process for inference can potentially be interpreted as top-down feedback disambiguating low level representations? Rao and Ballard's predictive coding/Kalman filter model 6 did integrate generafive feedback in the perceptual inference process, but it was primarily a linear model and thus severely limited in practical utility. The data-driven Markov Chain Monte Carlo approach of Zhu and colleagues 7, 8 might be the most successful recent application of this proposal in solving real and difficult computer vision problems using generafive models, though its connection to the visual cortex has not been explored. Here, we bring in a powerful and widely applicable paradigm from artificial intelligence and computer vision to propose some new ideas about the algorithms of visual cortical process- ing and the nature of representations in the visual cortex. We will review some of our and others' neurophysiological experimental data to lend support to these ideas

ing away from the specific coding of the spinal fields, the examples from neurobiology provide the framework for a motor control system based on a small number of additive primitives (or basis behaviors) sufficient for a rich output movement repertoire. Our previous work (Matari'c 1995, Matari'c 1997), inspired by the same biological results, has successfully applied the idea of basis behaviors to control of mobile robots 6 by fitting it directly into the modular behavior-based control paradigm. Applictions of schema theory (Arbib 1992) to behavior-based mobile robots (Arkin 1987) have employed a similar notion of composable behaviors, stemming from foundations in neuroscience (Arbib 1981, Arbib 1989). The idea of using such primitives for articulator control has been recently studied in robotics. Williamson (1996) and Marjanovi'c, Scassellati & Williamson (1996) developed a 6 DOF (degrees of freedom) robot arm controller. While in the biological and mobile robotics work primitives c...

A neural network model is developed to explain how visual thalamocortical interactions give rise to boundary percepts such as illusory contours and surface percepts such as filled-in brightnesses. Top-down feedback interactions are needed in addition to bottom-up feed-forward interactions to simulate these data. One feedback loop is modeled between lateral geniculate nucleus (LGN) and cortical area VI, and another within cortical areas VI and V2. The first feedback loop realizes a matching process which enhances LGN cell activities that are consistent with those of active cortical cells, and suppresses LGN activities that are not. This corticogeniculate feedback, being endstopped and oriented, also enhances LGN ON cell activations at the ends of thin dark lines, thereby leading to enhanced cortical brightness percepts when the lines group into closed illusory contours. The second feedback loop generates boundary representations, including illusory contours, that coherently bind distributed cortical features together. Brightness percepts form within the surface representations through a diffusive filling-in process that is contained by resistive gating signals from the boundary representations. The model is used to simulate illusory contours and surface brightnesses induced by Ehrenstein disks, Kanizsa squares, Glass patterns, and cafe wall patterns in single contrast, reverse contrast, and mixed contrast configurations. These examples illustrate how boundary

The processes whereby our brains continue to learn about a changing world in a stable fashion throughout life are proposed to lead to conscious experiences. These processes include the learning of top-down expectations, the matching of these expectations against bottom-up data, the focusing of attention upon the expected clusters of information, and the development of resonant states between bottom-up and top-down processes as they reach an attentive consensus between what is expected and what is there in the outside world. It is suggested that all conscious states in the brain are resonant states, and that these resonant states trigger learning of sensory and cognitive representations. The models which summarize these concepts are therefore called Adaptive Resonance Theory, or ART, models. Psychophysical and neurobiological data in support of ART are presented from early vision, visual object recognition, auditory streaming, variable-rate speech perception, somatosensory perception, a...

... This article models how these interactions help visual cortex to realize: (1) the binding process whereby cortex groups distributed data into coherent object representations; (2) the attentional process whereby cortex selectively processes important events; and (3) the developmental and learning processes whereby cortex shapes its circuits to match environmental constraints. New computational ideas about feedback systems suggest how neocortex develops and learns in a stable way, and why top-down attention requires converging bottom-up inputs to fully activate cortical cells, whereas perceptual groupings do not.

. A computational model of hippocampal activity during spatial cognition and navigation tasks is presented. The spatial representation in our model of the rat hippocampus is built on-line during exploration via two processing streams. An allothetic vision-based representation is built by unsupervised Hebbian learning extracting spatio-temporal properties of the environment from visual input. An idiothetic representation is learned based on internal movement-related information provided by path integration. On the level of the hippocampus, allothetic and idiothetic representations are integrated to yield a stable representation of the environment by a population of localized overlapping CA3-CA1 place fields. The hippocampal spatial representation is used as a basis for goal-oriented spatial behavior. We focus on the neural pathway connecting the hippocampus to the nucleus accumbens. Place cells drive a population of locomotor action neurons in the nucleus accumbens. Reward-based learnin...