Mark E. Oxley

The multilayer perceptron as an approximation to a Bayes optimal discriminant function

The multilayer perceptron, when trained as a classifier using backpropagation, is shown to approximate the Bayes optimal discriminant function. The result is demonstrated for both the two-class problem and multiple classes. It is shown that the outputs of the multilayer perceptron approximate the a posteriori probability functions of the classes being trained. The proof applies to any number of layers and any type of unit activation function, linear or nonlinear.

Physiologically motivated image fusion for object detection using a pulse coupled neural network

This paper presents the first physiologically motivated pulse coupled neural network (PCNN)-based image fusion network for object detection. Primate vision processing principles, such as expectation driven filtering, state dependent modulation, temporal synchronization, and multiple processing paths are applied to create a physiologically motivated image fusion network. PCNNs are used to fuse the results of several object detection techniques to improve object detection accuracy. Image processing techniques (wavelets, morphological, etc.) are used to extract target features and PCNNs are used to focus attention by segmenting and fusing the information. The object detection property of the resulting image fusion network is demonstrated on mammograms and Forward Looking Infrared Radar (FLIR) images. The network removed 94% of the false detections without removing any true detections in the FLIR images and removed 46% of the false detections while removing only 7% of the true detections in the mammograms. The model exceeded the accuracy obtained by any individual filtering methods or by logical ANDing the individual object detection technique results.

Neural networks for automatic target recognition

Abstract Many applications reported in artificial neural networks are associated with military problems. This paper reviews concepts associated with the processing of military data to find and recognize targets—automatic target recognition (ATR). A general-purpose automatic target recognition system does not exist. The work presented here is demonstrated on military data, but it can only be consideredproof of principle until systems are fielded andproven “under-fire”. ATR data can be in the form of non-imaging one-dimensional sensor returns, such as ultra-high range-resolution radar returns for air-to-air automatic target recognition and vibration signatures from a laser radar for recognition of ground targets. The ATR data can be two-dimensional images. The most common ATR images are infrared, but current systems must also deal with synthetic aperture radar images. Finally, the data can be three-dimensional, such as sequences of multiple exposures taken over time from a nonstationary world. Targets move, as do sensors, and that movement can be exploited by the ATR. Hyperspectral data, which are views of the same piece of the world looking at different spectral bands, is another example of multiple image data; the third dimension is now wavelength and not time. ATR system design usually consists of four stages. The first stage is to select the sensor or sensors to produce the target measurements. The next stage is the preprocessing of the data and the location of regions of interest within the data (segmentation). The human retina is a ruthless preprocessor. Physiology motivated preprocessing and segmentation is demonstrated along with supervised and unsupervised artificial neural segmentation techniques. The third design step is feature extraction and selection: the extraction of a set of numbers which characterize regions of the data. The last step is the processing of the features for decision making (classification). The area of classification is where most ATR related neural network research has been accomplished. The relation of neural classifiers to Bayesian techniques is emphasized along with the more recent use of feature sequences to enhance classification. The principal theme of this paper is that artificial neural networks have proven to be an interesting and useful alternate processing strategy. Artificial neural techniques, however, are not magical solutions with mystical abilities that work without good engineering. Good understanding of the capabilities and limitations of neural techniques is required to apply them productively to ATR problems.

Comparative analysis of backpropagation and the extended Kalman filter for training multilayer perceptrons

The relationship between backpropagation and extended Kalman filtering for training multilayer perceptrons is examined. These two techniques are compared theoretically and empirically using sensor imagery. Backpropagation is a technique from neural networks for assigning weights in a multilayer perceptron. An extended Kalman filter can also be used for this purpose. A brief review of the multilayer perceptron and these two training methods is provided. Then, it is shown that backpropagation is a degenerate form of the extended Kalman filter. The training rules are compared in two examples: an image classification problem using laser radar Doppler imagery and a target detection problem using absolute range images. In both examples, the backpropagation training algorithm is shown to be three orders of magnitude less costly than the extended Kalman filter algorithm in terms of a number of floating-point operations. >

Analytical modeling of aquifer decontamination by pumping when transport is affected by rate‐limited sorption

Aquifer cleanup efforts at contaminated sites frequently involve operation of a system of extraction wells. It has been found that contaminant load discharged by extraction wells typically declines with time, asymptotically approaching a residual level. Such behavior could be due to rate-limited desorption of an organic contaminant from aquifer solids. An analytical model is presented which accounts for rate-limited desorption of an organic solute during cleanup of a contaminated site. Model equations are presented which describe transport of a sorbing contaminant in a converging radial flow field, with sorption described by (1) equilibrium, (2) first-order rate, and (3) Fickian diffusion expressions. The model equations are solved in the Laplace domain and numerically inverted to simulate contaminant concentrations at an extraction well. A Laplace domain solution for the total contaminant mass remaining in the aquifer is also derived. It is shown that rate-limited sorption can have a significant impact upon aquifer remediation. Approximate equivalence among the various rate-limited models is also demonstrated.

Using explicit requirements and metrics for interface agent user model correction

The complexity of current computer systems and software vrarrants research into methods to decrease the cognitive load on users. Determining horr to get the right information into the right form vrith the right tool at the right time has bccomc a monumental task one necessitating intelligent intarfacc agents vlith the ability to predict the users’ needs and intent, An accurate user model is considered necessary for effective prediction of user intent. Methods for maintaining nccurato user models is the main thrust of this paper. We describe an approach for dynamically correcting an interface ngent’s user model based on utility theory. We explicitly take into account an agent’s requirements and metrics for mc,asuring the agent’s effectiveness of meeting those requirements, Using these requirements and metrics, me devclop a requirements utility function that determines when a user model should be corrected and how. We present a correction model based on a multi-agent bidding process and the aforementioned metrics and utility function. Finally, me discuss several critical research issues concerning the use of user models that open fertile ground for future research.

Object tracking through adaptive correlation

Current Air Force interests include a desire to track an object based on its shape once it has been designated as a target. The use of a correlation-based system to track an object through a series of images based on templates derived from previous image frames is discussed. The ability to track is extended to sequences that include multiple objects of interest within the field of view. This is accomplished by comparing the height and shape of the template autocorrelation to the peaks in the correlation of the template with the next scene. The result is to identify the region in the next scene that best matches the designated target. In addition to correlation plane postprocessing, an adaptive window is used to determine the template size to reduce the effects of correlator walkoff. The image sequences used were taken from a forward-looking infrared sensor mounted on board a DC-3 aircraft. The images contain a T-55 tank and both an M-113 and a TAB-71 armored personnel carrier moving in a columnized formation along a dirt road. This research presents techniques to (1) track targets in the presence of other, and sometimes brighter, targets of similar shape, (2) to maintain small tracking errors, and (3) to reduce the effects of correlator walk-off.

Communication Waveform Design Using an Adaptive Spectrally Modulated, Spectrally Encoded (SMSE) Framework

Fourth-generation (4G) communication systems will likely support multiple capabilities while providing universal, high-speed access. One potential enabler for these capabilities is software-defined radio (SDR). When controlled by cognitive radio (CR) principles, the required waveform diversity is achieved through a synergistic union called CR-based SDR. This paper introduces a general framework for analyzing, characterizing, and implementing spectrally modulated, spectrally encoded (SMSE) signals within CR-based SDR architectures. Given orthogonal frequency division multiplexing (OFDM) is one 4G candidate signal, OFDM-based signals are collectively classified as SMSE since data modulation and encoding are applied in the spectral domain. The proposed framework provides analytic commonality and unification of multiple SMSE signals. Framework applicability and flexibility is demonstrated for candidate 4G signals by: 1) showing that resultant analytic expressions are consistent with published results and 2) presenting representative modeling and simulation results to reinforce practical utility

Binary, ternary and quaternary compound former/nonformer prediction via Mendeleev number

A significant breakthrough has been achieved in the design of new materials by using materials databases, semiempirical approaches and neural networks. It was found in the present work that a nonlinear expression involving one elemental property parameter can be used to predict, with an overall accuracy exceeding 99%, the occurrence of a compound for any binary, ternary or quaternary system. This elemental property parameter, referred to as the Mendeleev number, was conceived by D.G. Pettifor in 1983 to group binary compounds by crystal structures. The immediate profit of this discovery is the obvious savings, in time and resources, relative to the investigation of yet-to-be-studied, materials systems. In the longer term the relation found here will make it possible to better define the search space for the development of new materials and encourage attempts to predict more specific information such as stoichiometries, crystal structures and physical properties.

Cohort selection and word grammar effects for speaker recognition

Automatic speaker recognition systems are maturing and databases have been designed to specifically compare algorithms and results to target error rates. The LDC YOHO speaker verification database was designed to test error rates at the 1% false rejection and 0.1% false acceptance level. This work examines the use of speaker-dependent (SD) monophone models to meet these requirements. By representing each speaker with 22 monophones, both closed-set speaker identification and global-threshold verification was performed. Using four combination lock phrases, speaker identification error rates are obtained at 0.19% for males and 0.31% for females. By defining a test hypothesis, a critical error analysis for speaker verification is developed and new results reported for YOHO. A new Bhattacharyya distance is developed for cohort selection. This method, based on the second order statistics of the enrolment Viterbi log-likelihoods, determines the optimal cohorts and achieves an equal error rate of 0.282%.