We present a neural network algorithm that simultaneously performs seg(cid:173) mentation and recognition of input patterns that self-organizes to detect input pattern locations and pattern boundaries. We demonstrate this neu(cid:173) ral network architecture on character recognition using the NIST database and report on results herein. The resulting system simultaneously seg(cid:173) ments and recognizes touching or overlapping characters, broken charac(cid:173) ters, and noisy images with high accuracy.

A CCD-based processor that we call the NNC2 is presented. The NNC2 implements a fully connected 192-input, 32-output two-layer network and can be cascaded to form multilayer networks or used in parallel for ad(cid:173) ditional input or output nodes. The device computes 1.92 x 109 connec(cid:173) tions/sec when clocked at 10 MHz. Network weights can be specified to six bits of accuracy and are stored on-chip in programmable digital memories. A neural network pattern recognition system using NNC2 and CCD im(cid:173) age feature extractor (IFE) devices is described.

A combined neural network and rule-based approach is suggested as a general framework for pattern recognition. This approach enables unsu(cid:173) pervised and supervised learning, respectively, while providing probability estimates for the output classes. The probability maps are utilized for higher level analysis such as a feedback for smoothing over the output la(cid:173) bel maps and the identification of unknown patterns (pattern "discovery"). The suggested approach is presented and demonstrated in the texture - analysis task. A correct classification rate in the 90 percentile is achieved for both unstructured and structured natural texture mosaics.

We compare two strategies for training connectionist (as well as non(cid:173) connectionist) models for statistical pattern recognition. The probabilistic strat(cid:173) egy is based on the notion that Bayesian discrimination (i.e .• The differential strategy is based on the notion that the identity of the largest class a posteriori probability of the feature vector is all that is needed to achieve Bayesian discrimination. Each strategy is directly linked to a family of objective functions that can be used in the supervised training procedure. We prove that the probabilistic strategy - linked with error measure objective functions such as mean-squared-error and cross-entropy - typically used to train classifiers necessarily requires larger training sets and more complex classifier architectures than those needed to approximate the Bayesian discrim(cid:173) linked inant function.

We have investigated the properties of neurons in inferior temporal (IT) cortex in monkeys performing a pattern matching task. Simple back(cid:173) propagation networks were trained to discriminate the various stimulus conditions on the basis of the measured neuronal signal. We also trained networks to predict the neuronal response waveforms from the spatial pat(cid:173) terns of the stimuli. The results indicate t.hat IT neurons convey tempo(cid:173) rally encoded information about both current and remembered patterns, as well as about their behavioral context.

A new classifier is presented for text-independent speaker recognition. The new classifier is called the modified neural tree network (MNTN). The NTN is a hierarchical classifier that combines the properties of decision trees and feed-forward neural networks. The MNTN differs from the stan(cid:173) dard NTN in that a new learning rule based on discriminant learning is used, which minimizes the classification error as opposed to a norm of the approximation error. The MNTN also uses leaf probability mea(cid:173) sures in addition to the class labels.

The efficiency of image search can be greatly improved by using a coarse-to-fine search strategy with a multi-resolution image representa(cid:173) tion. However, if the resolution is so low that the objects have few dis(cid:173) tinguishing features, search becomes difficult. We show that the performance of search at such low resolutions can be improved by using context information, i.e., objects visible at low-resolution which are not the objects of interest but are associated with them. The networks can be given explicit context information as inputs, or they can learn to detect the context objects, in which case the user does not have to be aware of their existence. We also use Integrated Feature Pyramids, which repre(cid:173) sent high-frequency information at low resolutions.

In remote sensing applications "ground-truth" data is often used as the basis for training pattern recognition algorithms to gener(cid:173) ate thematic maps or to detect objects of interest. In practical situations, experts may visually examine the images and provide a subjective noisy estimate of the truth. Calibrating the reliability and bias of expert labellers is a non-trivial problem. In this paper we discuss some of our recent work on this topic in the context of detecting small volcanoes in Magellan SAR images of Venus. Empirical results (using the Expectation-Maximization procedure) suggest that accounting for subjective noise can be quite signifi(cid:173) cant in terms of quantifying both human and algorithm detection performance.

We have developed a foveated gesture recognition system that runs in an unconstrained office environment with an active camera. Us(cid:173) ing vision routines previously implemented for an interactive envi(cid:173) ronment, we determine the spatial location of salient body parts of a user and guide an active camera to obtain images of gestures or expressions. A hidden-state reinforcement learning paradigm is used to implement visual attention. The attention module selects targets to foveate based on the goal of successful recognition, and uses a new multiple-model Q-Iearning formulation. Given a set of target and distractor gestures, our system can learn where to foveate to maximally discriminate a particular gesture.

Completely parallel object recognition is NP-complete. Achieving a recognizer with feasible complexity requires a compromise be(cid:173) tween parallel and sequential processing where a system selectively focuses on parts of a given image, one after another. Successive fixations are generated to sample the image and these samples are processed and abstracted to generate a temporal context in which results are integrated over time. A computational model based on a partially recurrent feedforward network is proposed and made cred(cid:173) ible by testing on the real-world problem of recognition of hand(cid:173) written digits with encouraging results.