Computer Sciences Commons^™

Text Independent Speaker Verification Using Binary-Pair Partitioned Neural Networks, Claude A. Norton Iii

Electrical & Computer Engineering Theses & Dissertations

A method is presented for the application of binary-pair partitioned neural networks to the task of speaker verification. This technique is based on a previously developed neural network classifier for speaker identification.

The main focus of this research was the development and testing of the algorithms necessary to extend the binary-pair partitioning approach from speaker identification to speaker verification. The method is based on the development of a user profile which is obtained from discriminative data provided by the binary-pair partitioned neural networks.

Experimental results are provided which demonstrate the viability of this approach, using the TIMIT speech corpus for …

Atomic Broadcast In Heterogeneous Distributed Systems, Osman Zeineldine

Computer Science Theses & Dissertations

Communication services have long been recognized as possessing a dominant effect on both performance and robustness of distributed systems. Distributed applications rely on a multitude of protocols for the support of these services. Of crucial importance are multicast protocols. Reliable multicast protocols enhance the efficiency and robustness of distributed systems. Numerous reliable multicast protocols have been proposed, each differing in the set of assumptions adopted, especially for the communication network. These assumptions make each protocol suitable for a specific environment. The presence of different distributed applications that run on different LANs and single distributed applications that span different LANs mandate …

Safe Stratified Datalog With Integer Order Programs, Peter Revesz

School of Computing: Conference and Workshop Papers

Guaranteeing termination of programs on all valid inputs is important for database applications. Termination cannot be guaranteed in Stratified Datalog with integer (gap)-order programs on generalized databases because they express any Turing-computable function. This paper introduces a restriction of those programs that can express only computable queries. The restricted language has a high expressive power and a non-elementary data complexity.

Comparison Of System Identification Techniques For The Dexterous Orbital Servicing System (Doss), Allan Terrence Morris

Electrical & Computer Engineering Theses & Dissertations

The major contribution of system identification is the development of analytical models of a system. Many identification techniques can adequately describe the input and output map. One of the challenges of these techniques is to correctly infer from the measured data the characteristics of the individual, contributing components producing a more accurate system model.

In this thesis linear, dynamic, multivariable state-space models for three joints of the Dexterous Orbital Servicing System (DOSS) are identified. DOSS is a representative space station manipulator at the NASA Langley Research Center. The dynamic models of the manipulator will first be estimated by applying nonparametric …

Optimal Mixed-Norm Control Synthesis For Discrete-Time Linear Systems, David R. Jacques

Theses and Dissertations

A mixed-norm approach to control synthesis for discrete time linear systems is developed. Specifically, the problem of minimizing the H₂ norm of a transfer function, subject to a combination of ℓ₁ and-or H_∞ norm constraints on dissimilar but related transfer functions is considered. The uniqueness of the optimal solution is shown, and numerical methods for approximating the optimal solution to within arbitrary accuracy are developed. These methods generally result in high order compensators which can not be implemented in most practical applications. In response to this, a numerical method is developed which solves for suboptimal solutions of …

Electric Field Mapping System With Nanosecond Temporal Rosolution, F. E. Peterkin, R. Block, K. H. Schoenbach

Bioelectrics Publications

The electric field dependence of the absorption coefficient in semi‐insulating GaAs at the absorption edge was measured in a high‐voltage pulsed experiment. Pulse duration was kept below 50 ns in order to avoid thermal effects. A GaAs laser diode was used as a probe light source with wavelength varied from 902 to 911 nm. For fields up to 40 kV/cm the absorption coefficient increased from 3 to 17 cm⁻¹ at 902 nm, with smaller absolute increases evident at the longer wavelengths. Calculation from theory was consistent with this behavior. The spatial variation of the electric field was also recorded …

Analysis And Extension Of Model-Based Software Executives, Keith E. Lewis

Theses and Dissertations

This research developed a comprehensive description of the simulation environment of Architect, a domain-oriented application composition system being developed at the Air Force Institute of Technology to explore new software engineering technologies. The description combines information from several previous research efforts and Architect's source code into a single, comprehensive document. A critical evaluation of the simulation environment was also performed, identifying improvements and modifications that enhance Architecture's application execution capabilities by reducing complexity and execution time. The analysis was then taken one step further and presented extensions to the current simulation environment. The extensions included investigating the feasibility of mixed-mode …

Hardware Assists For High Performance Computing Using A Mathematics Of Arrays, Hardy J. Pottinger, W. Eatherton, J. Kelly, T. Schiefelbein, Lenore Mullin, R. Ziegler

Electrical and Computer Engineering Faculty Research & Creative Works

Work in progress at the University of Missouri-Rolla on hardware assists for high performance computing is presented. This research consists of a novel field programmable gate array (FPGA) based reconfigurable coprocessor board (the Chameleon Coprocessor) being used to evaluate hardware architectures for speedup of array computation algorithms. These algorithms are developed using a Mathematics of Arrays (MOA). They provide a means to generate addresses for data transfers that require less data movement than more traditional algorithms. In this manner, the address generation algorithms are acting as an intelligent data prefetching mechanism or special purpose cache controller. Software implementations have been …

Next Generation Real-Time Systems: Investigating The Potential Of Partial-Solution Tasks, Robert E. J. Caley

Theses and Dissertations

While the cyclic executive and fixed-priority scheduling strategies have been sufficient to handle traditional real- time requirements. they are insufficient for dealing with the complexities of next-generation real-time systems. New methods of intelligent control must be developed for guaranteeing on-time task completion for real-time systems that are faced with unpredictable and dynamically changing requirements. Implementing real-time processes as partial-solution tasks is one technique that may be beneficial. This type of task. when combined with intelligent control, has the potential for increasing pre-runtime schedulability, system maintainability. and runtime robustness. This research investigates the benefits of partial-solution tasks by experimentally measuring the …

Survey And Implementation Of Commercial Manual Controllers For A Generic Telerobotics Architecture, Thomas E. Deeter

Theses and Dissertations

The purpose of this study is to determine an input device for the Air Force's generic telerobotics architecture for large aircraft maintenance and repair. One area of concern is the human to machine interface, more specifically, which manual controller should be used for the specified tasks in this architecture. We mailed a survey to 68 companies in order to compile a list of possible input devices that the telerobotics architecture could use. 32 companies responded which gave me enough data to generate a list that described the physical traits of the input devices. We then divided the required tasks into …

Embedology And Neural Estimation For Time Series Prediction, Robert E. Garza

Theses and Dissertations

Time series prediction has widespread application, ranging from predicting the stock market to trying to predict future locations of scud missiles. Recent work by Sauer and Casdagli has developed into the embedology theorem, which sets forth the procedures for state space manipulation and reconstruction for time series prediction. This includes embedding the time series into a higher dimensional space in order to form an attractor, a structure defined by the embedded vectors. Embedology is combined with neural technologies in an effort to create a more accurate prediction algorithm. These algorithms consist of embedology, neural networks, Euclidean space nearest neighbors, and …

Image Perception Wavelet Simulation And Enhancement For The Visually Impaired, Lemuel Ray Myers Jr.

Theses and Dissertations

This research delves into the area of image enhancement for the visually impaired. Binocular macular degeneration a visual impairment, affects many Americans; since this condition could not be corrected with conventional glasses the literature suggested using an enhancement system which used a pre-emphasis algorithm to enhance the input image for output to the observer. The work of Dr. Eli Peli, a pioneer in the field of image enhancement, is examine and reproduced. Since his work concentrated mainly on frequency analysis of images, the bulk of this research involves using discrete wavelet analysis to augment that work. A biorthogonal wavelet set …

The Fat-Pyramid And Universal Parallel Computation Independent Of Wire Delay, Ronald I. Greenberg

Computer Science: Faculty Publications and Other Works

This paper shows that a fat-pyramid of area Θ(A) requires only O(log A) slowdown to simulate any competing network of area A under very general conditions. The result holds regardless of the processor size (amount of attached memory) and number of processors in the competing networks as long as the limitation on total area is met. Furthermore, the result is valid regardless of the relationship between wire length and wire delay. We especially focus on elimination of the common simplifying assumption that unit time suffices to traverse a wire regardless of its length, since the assumption becomes more and more …

Biofeedback Gait Training Auditory Vs. Visual Techniques, Timothy D. Hiemenz

Electrical & Computer Engineering Theses & Dissertations

Training lower extremity amputees to walk normally is quite a difficult task. Amputees must wear a prosthesis so they can walk at all. This allows them some mobility, but their walking pattern may be unnatural. If their gait is temporally asymmetric, they need to exert more energy to move about. This research was initiated to help lower extremity amputees to walk more efficiently using biofeedback gait training. Two types of feedback were developed and tested to determine which method gave the most understandable feedback, validating its use in a clinical setting.

A normal gait cycle uses the lower limbs to …

Hardware Assists For High Performance Computing Using A Mathematics Of Arrays, H. Pottinger, W. Eatherton, L. Mullin, T. Schiefelbein

Computer Science Technical Reports

Work in progress at the University of Missouri - Rolla on hardware assists for high performance computing and communication is presented. This research consists of a field programmable gate array based rapid prototyping environment (Chameleon) being used to evaluate hardware architectures for speedup of array partitioning and routing algorithms. These algorithms have been developed using a Mathematics of Arrays (MOA) and are optimal in the sense that they reduce normal array indexing operations to a series of primitive additions based on array shape. The method is simple yet powerful, produces algorithms which are provably correct, and are independent of dimensionality. …

An Fpga Based Reconfigurable Coprocessor Board Utilizing A Mathematics Of Arrays, H. Pottinger, W. Eatherton, T. Schiefelbein, L. Mullin, R. Ziegler

Computer Science Technical Reports

Work in progress at the University of Missouri-Rolla on hardware assists for high performance computing is presented. This research consists of a novel field programmable gate array (FPGA) based reconfigurable coprocessor board (the Chameleon Coprocessor) being used to evaluate hardware architectures for speedup of array computation algorithms. These algorithms have been developed using a Mathematics of Arrays (MOA) and are optimal in the sense that they reduce normal array cartesian indexing operations to a series of primitive additions and offsets based on array shape. The method is simple yet powerful, produces algorithms which are provably correct, and are independent of …

Multirate Time-Frequency Distributions, John R. O'Hair

Theses and Dissertations

Multirate systems, which find application in the design and analysis of filter banks, are demonstrated to also be useful as a computational paradigm. It is shown that any problem which can be expressed a set of vector-vector, matrix-vector or matrix-matrix operations can be recast using multirate. This means all of numerical linear algebra can be recast using multirate as the underlying computational paradigm. As a non-trivial example, the multirate computational paradigm is applied to the problem of Generalized Discrete Time- Frequency Distributions GDTFD to create a new family of fast algorithms. The first of this new class of distributions is …

Constraint Objects, Divesh Srivastava, Raghu Ramakrishnan, Peter Revesz

School of Computing: Conference and Workshop Papers

We describe the Constraint Object Data Model (CODM), which enhances an object-based data model with existential constraints to naturally represent partially specified information. We present the Constraint Object Query Language (COQL), a declarative, rule-based query language that can be used to infer relationships about and monotonically refine information represented in the CODM. COQL has a model-theoretic and an equivalent fixed-point semantics, based on the notions of constraint entailment and "proofs in all possible worlds." We also provide a novel polynomial-time algorithm for quantifier elimination for set-order constraints, a restricted class of set constraints that uses membership of subset-equal.

Dynamic Task Scheduling For The Atamm Multicomputer Operating System Using Embedded Firmware On Microcontrollers, Sudhir Sastry

Electrical & Computer Engineering Theses & Dissertations

A dynamic task scheduling strategy for the distributed processing of large grain dataflow algorithms using embedded firmware on an ATAMM testbed consisting of interconnected microcontrollers is presented in this thesis. The ODU/NASA developed Algorithm to Architecture Mapping Model, ATAMM, uses marked graph models to specify data and control flow for the execution of iterative, deterministic large grain dataflow algorithms in a multicomputing environment. The testbed consists of a bank of four 68HC11 microcontrollers that communicate over a token bus. The token bus arbitration scheme used is contention free and well suited for real-time computing applications. The execution of data flow …

Clustering Techniques In Speaker Recognition, Douglas N. Prescott

Theses and Dissertations

This thesis presents a comparison based on identification rate, of three clustering techniques applied to cepstral features for speaker identification. LBG vector quantization as developed by Linde, Buzo and Gray; is used to provide benchmark performance for comparison with Fuzzy clustering (based on the unsupervised fuzzy partition-optimal number of classes, UFP-ONC algorithm by Gath and Geva) and an Artificial Neural Network, the Multilayer Perceptron. Cepstral features from the TIMIT, King and AFIT93 corpus speaker databases are used to produce speaker-identification classifiers using each of the clustering algorithms. The experiment reported evaluates the speaker identification performance using the 20-dimensional cepstral features …

Scheduling Of Unstructured Communication On The Intel Ipsc/860, Jhy-Chun Wang, Sanjay Ranka

College of Engineering and Computer Science - Former Departments, Centers, Institutes and Projects

In this paper we present several algorithms for decomposing all-to-many personalized communication into a set of disjoint partial permutations. These partial permutations avoid node contention as well as link contention. We discuss the theoretical complexity of these algorithms and study their effectiveness both from the view of static scheduling and from runtime scheduling. Experimental results for our algorithms are presented on the iPSC/860.

Neural Network Diagnosis Of Malignant Melanoma From Color Images, Fikret Erçal, Hsi-Chieh Lee, William V. Stoecker, Randy Hays Moss, Anurag Chawla

Computer Science Faculty Research & Creative Works

Malignant melanoma is the deadliest form of all skin cancers. Approximately 32,000 new cases of malignant melanoma were diagnosed in 1991 in the United States, with approximately 80% of patients expected to survive 5 years. Fortunately, if detected early, even malignant melanoma may be treated successfully, Thus, in recent years, there has been rising interest in the automated detection and diagnosis of skin cancer, particularly malignant melanoma. Here, the authors present a novel neural network approach for the automated separation of melanoma from 3 benign categories of tumors which exhibit melanoma-like characteristics. The approach uses discriminant features, based on tumor …

Filtering, Coding, And Compression With Malvar Wavelets, Stephen R. Hall

Theses and Dissertations

This thesis develops and evaluates a number of new concepts and tools for the analysis of signals using Malvar wavelets lapped orthogonal transforms . Windowing, often employed as a spectral estimation technique, can result in irreparable distortions in the transformed signal. By utilizing the Malvar wavelet, any signal distortion resulting from the transformation can be eliminated or cancelled during reconstruction. This is accomplished by placing conditions on the window and the basis function and then incorporating the window into the orthonormal representation. Paradigms for both a complex-valued and a real-valued Malvar wavelet are summarized. I-lie algorithms for the wavelets were …

Developing A Graphical User Interface To Support A Real-Time Digital Signal Processing System, Jeffrey C. Miller

Theses and Dissertations

A graphical software user interface for a VMEbus-based real-time digital signal processing system was designed. Use, requirements were defined and the Rumbaugh object-oriented analysis and design technique was applied to analyze the requirements and produce an object-oriented design. The software design includes a graphical, mouse- and keyboard driven user interface, specialized hardware driver modules, and operating system interfaces. An implementation plan was also developed to map the design into the C programming language using existing system code, automatically generated code, and newly written code. Based on the implementation plan, a limited software system prototype w successfully developed and demonstrated. The …

Investigation Into Model-Based Fuzzy Logic Control, Michael W. Logan

Theses and Dissertations

This thesis investigates the feasibility of a proposed hybrid linear/Fuzzy controller for nonlinear plants. The proposed controller concept is based on the use of multiple linearizations of a nonlinear plant, which describe the dynamics of perturbations about equilibrium points throughout the desired envelope of operation. A bank of linear compensators is developed, each corresponding to a linearized plant about a different equilibrium. The multiple control signals generated by the bank of compensators are then weighted and summed using Fuzzy Logic to produce a composite control perturbation signal, which is used to drive the nonlinear plant. Experiments were conducted to test …

Parallel Algorithms For Single-Layer Channel Routing, Ronald I. Greenberg, Shih-Chuan Hung, Jau-Der Shih

Computer Science: Faculty Publications and Other Works

We provide efficient parallel algorithms for the minimum separation, offset range, and optimal offset problems for single-layer channel routing. We consider all the variations of these problems that have linear-time sequential solutions rather than limiting attention to the ``river-routing'' context, where single-sided connections are disallowed. For the minimum separation problem, we obtain O(lgN) time on a CREW PRAM or O(lgN/lglgN) time on a CRCW PRAM, both with optimal work (processor-time product) of O(N), where N is the number of terminals. For the offset range problem, we obtain the same time and processor bounds as long as only one side of …

Feasible Offset And Optimal Offset For Single-Layer Channel Routing, Ronald I. Greenberg, Jau-Der Shih

Computer Science: Faculty Publications and Other Works

The paper provides an efficient method to find all feasible offsets for a given separation in a VLSI channel routing problem in one layer. The prior literature considers this task only for problems with no single-sided nets. When single-sided nets are included, the worst-case solution time increases from Theta(n) to Omega(n^2), where n is the number of nets. But, if the number of columns c is O(n), one can solve the problem in time O(n^{1.5}lg n ), which improves upon a `naive' O(cn) approach. As a corollary of this result, the same time bound suffices to find the optimal offset …

A Performance Prediction Model For A Fault-Tolerant Computer During Recovery And Restoration, Rodrigo A. Obando

Electrical & Computer Engineering Theses & Dissertations

The modeling and design of a fault-tolerant multiprocessor system is addressed in this dissertation. In particular, the behavior of the system during recovery and restoration after a fault has occurred is investigated. Given that a multicomputer system is designed using the Algorithm to Architecture To Mapping Model (ATAMM) model, and that a fault (death of a computing resource) occurs during its normal steady-state operation, a model is presented as a viable research tool for predicting the performance bounds of the system during its recovery and restoration phases. Furthermore, the bounds of the performance behavior of the system during this transient …

Recognition Of Quadric Surfaces From Range Data: An Analytical Approach, Ivan X. D. D'Cunha

Electrical & Computer Engineering Theses & Dissertations

In this dissertation, a new technique based on analytic geometry for the recognition and description of three-dimensional quadric surfaces from range images is presented. Beginning with the explicit representation of quadrics, a set of ten coefficients are determined for various three-dimensional surfaces. For each quadric surface, a unique set of two-dimensional curves which serve as a feature set is obtained from the various angles at which the object is intersected with a plane. Based on a discriminant method, each of the curves is classified as a parabola, circle, ellipse, hyperbola, or a line. Each quadric surface is shown to be …

Middle Terminal Cell Models For Efficient Over-The-Cell Routing, Siddharth Bhingarde

Masters Theses

In this thesis, we introduce a new class of cell models called Middle Terminal Models (MTM) which combines the advantages of existing cell models: BTM and CTM. This class contains the models 2MTM, 3MTM-V, and 3MTM+V depending on the number of metal layers and the permissibility of vias in over-the-cell areas. In MTM, two rows of terminals are located in the middle of the cell. This partitions over-the-cell area into three regions and allows allocation of more nets to over-the-cell area in congested channels. We prove that when vias are allowed over-the-cell, in "almost all" cases the MTM based layouts …