PERCEPTUAL systems routinely separate content from style, classifying familiar words spoken in an unfamiliar accent, identifying a font or handwriting style across letters, or recognizing a familiar face or object seen under unfamiliar viewing conditions. Yet a general and tractable computational model of this ability to untangle the underlying factors of perceptual observations remains elusive. Existing factor models are either insufficiently rich to capture the complex interactions of perceptually meaningful factors such as phoneme and speaker accent or letter and font, or do not allow efficient learning algorithms. Here we show how perceptual systems may learn to solve these crucial tasks using surprisingly simple bilinear models. We report promising results in three realistic perceptual domains: spoken vowel classification with a benchmark multi-speaker database, extrapolation of fonts to unseen letters, and translation of faces to novel illuminants.

this paper, we describe a hierarchical network model of visual recognition that explains these experimental observations by using a form of the extended Kalman filter as given by the Minimum Description Length (MDL) principle. The model dynamically combines input-driven bottom-up signals with expectation-driven top-down signals to predict current recognition state. Synaptic weights in the model are adapted in a Hebbian manner according to a learning rule also derived from the MDL principle. The resulting prediction/learning scheme can be viewed as implementing a form of the Expectation-Maximization (EM) algorithm. The architecture of the model posits an active computational role for the reciprocal connections between adjoining visual cortical areas in determining neural response properties. In particular, the model demonstrates the possible role of feedback from higher cortical areas in mediating neurophysiological effects due to stimuli from beyond the classical receptive field. Si

We present a general encoding-decoding framework for interpreting the activity of a population of units. A standard population code interpretation method, the Poisson model, starts from a description as to how a single value of an underlying quantity can generate the activities of each unit in the population. In casting it in the encoding-decoding framework, we find that this model is too restrictive to describe fully the activities of units in population codes in higher processing areas, such as the medial temporal area. Under a more powerful model, the population activity can convey information not only about a single value of some quantity but also about its whole distribution, including its variance, and perhaps even the certainty the system has in the actual presence in the world of the entity generating this quantity. We propose a novel method for forming such probabilistic interpretations of population codes and compare it to the existing method.

Sensorimotor transformations are nonlinear mappings of sensory inputs to motor responses. We explore here the possibility that the responses of single neurons in the parietal cortex serve as basis functions for these transformations. Basis function decomposition is a general method for approximating nonlinear functions that is computationally efficient and well suited for adaptive modification. In particular, the responses of single parietal neurons can be approximated by the product of a Gaussian function of retinal location and a sigmoid function of eye position, called a gain field. A large set of such functions forms a basis set that can be used to perform an arbitrary motor response through a direct projection. We com-pare this hypothesis with other approaches that are commonly used to model population codes, such as computational maps and vectorial representations. Neither of these alternatives can fully account for the responses of parietal neurons, and they are computationally less efficient for nonlinear transformations. Basis functions also have the advantage of not depending on any coordinate system or reference frame. As a consequence, the position of an object can be represented in multiple reference frames simultaneously, a property consistent with the behavior of hemineglect patients with lesions in the parietal cortex.

Abstract not found

Using experimental facts about long-term potentiation (LTP) and hippocampal place cells, we model how a spatial map of the environment can be created in the rat hippocampus. Sequential firing of place cells during exploration induces, in the model, a pattern of LTP between place cells that shifts the location coded by their ensemble activity away from the actual location of the animal. These shifts provide a navigational map that, in a simulation of the Morris maze, can guide the animal toward its goal. The model demonstrates how behaviorally generated modifications of synaptic strengths can be read out to affect subsequent behavior. Our results also suggest a way that navigational maps can be constructed from experimental recordings of hippocampal place cells. *Current address: Dept. of Brain and Cognitive Sciences, MIT E25-236, 45 Carlton St., Cambridge, MA 02139. Blockade of long term potentiation (LTP) and hippocampal lesions drastically impair

As people examine their world, the proximal stimulus changes position on their retinae with every saccade, but they perceive the world as being stable. This phenomenon of visual stability was explored by making changes in natural, full-color pictures during selected saccades as observers examined them in preparation for a recognition test. In Experiment 1, the pictures were displaced up, down, left, or right by 0.3, 0.6, or 1.2°. In Experiment 2, the pictures were expanded or contracted by 10 % or 20%. As a secondary task, subjects pressed a button when a change was detected. Three results from previous studies with simpler stimuli did not generalize. Evidence suggests that subjects ' detection of image changes primarily involves the use of local information in the region of the eyes ' landing position. A saccade target theory of visual stability is proposed. Making a saccadic eye movement causes a displacement of the light pattern across the retinae. If a similar retinal displacement occurs during an eye fixation, there is perception of movement, that is, the world appears to jump. However, the same pattern of motion on the retinae, occurring as a consequence of making a saccade, is not perceived and the world appears stable. 1 This phenomenon, referred to traditionally as space constancy, and which we will call visual stability, permits people to visually explore the world with a moving sensory matrix without misattributing selfinduced stimulus motion on the matrix to the world itself. How the visual system achieves this stability has been a matter of speculation and research since Helmholtz (18667 1963) discussed the problem.

We analyze a model of navigational map formation based on correlation-based, temporally asymmetric potentiation and depression of synapses between hippocampal place cells. We show that synaptic modification during random exploration of an environment shifts the location encoded by place cell activity in such a way that it indicates the direction from any location to a fixed target avoiding walls and other obstacles. Multiple maps to different targets can be simultaneously stored if we introduce target-dependent modulation of place cell activity. Once maps to a number of target locations in a given environment have been stored, novel maps to previously unknown target locations are automatically constructed by interpolation between existing maps.

Population coding, where neurons with broad and overlapping firing rate tuning curves collectively encode information about a stimulus, is a common feature of sensory systems.We use decoding methods and measured properties of NMDA-mediated LTP induction to study the impact of long-term potentiation of synapses between the neurons of such a coding array. We find that, due to a temporal asymmetry in the induction of NMDA-mediated LTP, firing patterns in a neuronal array that initially represent the current value of a sensory input will, after training, provide an experienced-based prediction of that input instead. We compute how this prediction arises from and depends on the training experience. We also show how the encoded prediction can be used to generate learned motor sequences, such as the movement of a limb. This involves a novel form of memory recall that is driven by the motor response so that it automatically generates new information at a rate appropriate for the task being per...

Abstract not found