The primary goal of pattern recognition is supervised or unsupervised classification. Among the various frameworks in which pattern recognition has been traditionally formulated, the statistical approach has been most intensively studied and used in practice. More recently, neural network techniques and methods imported from statistical learning theory have been receiving increasing attention. The design of a recognition system requires careful attention to the following issues: definition of pattern classes, sensing environment, pattern representation, feature extraction and selection, cluster analysis, classifier design and learning, selection of training and test samples, and performance evaluation. In spite of almost 50 years of research and development in this field, the general problem of recognizing complex patterns with arbitrary orientation, location, and scale remains unsolved. New and emerging applications, such as data mining, web searching, retrieval of multimedia data, face recognition, and cursive handwriting recognition, require robust and efficient pattern recognition techniques. The objective of this review paper is to summarize and compare some of the well-known methods used in various stages of a pattern recognition system and identify research topics and applications which are at the forefront of this exciting and challenging field.

/pdf/statistical-pattern-recognition-a-review-1kz1b0tte4.pdf

Statistical pattern recognition: a review

A computational model is described in which the sizes of variables are represented by the explicit times at which action potentials occur, rather than by the more usual 'firing rate' of neurons. The comparison of patterns over sets of analogue variables is done by a network using different delays for different information paths. This mode of computation explains how one scheme of neuroarchitecture can be used for very different sensory modalities and seemingly different computations. The oscillations and anatomy of the mammalian olfactory systems have a simple interpretation in terms of this representation, and relate to processing in the auditory system. Single-electrode recording would not detect such neural computing. Recognition 'units' in this style respond more like radial basis function units than elementary sigmoid units.

Pattern recognition computation using action potential timing for stimulus representation

Image processing with neural networks–a review

Cyclostationarity: half a century of research

Classical feature extraction and data projection methods have been well studied in the pattern recognition and exploratory data analysis literature. We propose a number of networks and learning algorithms which provide new or alternative tools for feature extraction and data projection. These networks include a network (SAMANN) for J.W. Sammon's (1969) nonlinear projection, a linear discriminant analysis (LDA) network, a nonlinear discriminant analysis (NDA) network, and a network for nonlinear projection (NP-SOM) based on Kohonen's self-organizing map. A common attribute of these networks is that they all employ adaptive learning algorithms which makes them suitable in some environments where the distribution of patterns in feature space changes with respect to time. The availability of these networks also facilitates hardware implementation of well-known classical feature extraction and projection approaches. Moreover, the SAMANN network offers the generalization ability of projecting new data, which is not present in the original Sammon's projection algorithm; the NDA method and NP-SOM network provide new powerful approaches for visualizing high dimensional data. We evaluate five representative neural networks for feature extraction and data projection based on a visual judgement of the two-dimensional projection maps and three quantitative criteria on eight data sets with various properties. >

Artificial neural networks for feature extraction and multivariate data projection

A neural model approach to perform adaptive calculation of the principal components (eigenvectors) of the covariance matrix of an input sequence is proposed. The algorithm is based on the successive application of the modified Hebbian learning rule proposed by Oja on every new covariance matrix that results after calculating the previous eigenvectors. The approach is shown to converge to the next dominant component that is linearly independent of all previously determined eigenvectors. The optimal learning rate is calculated by minimising an error function of the learning rate along the gradient descent direction. The approach is applied to encode grey-level images adaptively, by calculating a limited number of the KLT coefficients that meet a specified performance criterion. The effect of changing the size of the input sequence (number of image subimages), the maximum number of coding coefficients on the bit-rate values, the compression ratio, the signal-to-noise ratio, and the generalisation capability of the model to encode new images are investigated.

Neural model for Karhunen-Loeve transform with application to adaptive image compression

Vector quantization (VQ) and block truncation coding (BTC) are successful image compression techniques. However, a reproduced image using VQ or BTC suffers from edge degradation. A new technique that combines the advantages of both VQ and BTC to combat this degradation is presented and is referred to as VQ-BTC. In VQ-BTC, a low-detail block is encoded using VQ. For a high-detail block, a modification of BTC is used to determine the locations of the relatively lighter and relatively darker pixels inside the block and VQ is then used to encode each. VQ-BTC provides improved edge reproduction and much lower bit rates than those obtained by BTC. >

Image compression using VQ-BTC

Neural networks for maximum likelihood clustering

A design technique for variable two-dimensional recursive digital filters

A neural model approach to adaptively calculating the principal components of the covariance matrix of an input sequence is proposed. The algorithm is based on the successive application of the modified Hebbian learning rule proposed by E. Oja (1982) on every covariance matrix which results after calculating the previous eigenvectors. This is equivalent to removing one dimension of the orthogonal space in which the data could be represented. Adopting a modification rule for the learning rate achieves faster convergence than that obtained when using other models. The optimal learning rate is calculated by minimizing an error function of the learning rate along the gradient descent direction. >

M.M. Fahmy

Papers

Neural model for Karhunen-Loeve transform with application to adaptive image compression

Image compression using VQ-BTC

Neural networks for maximum likelihood clustering

A design technique for variable two-dimensional recursive digital filters

A neural model for adaptive Karhunen Loeve transformation (KLT)