Digital Commons Network^™

Error Correction Via A Post-Processor For Continuous Speech Recognition, Eric K. Ringger, James F. Allen

Faculty Publications

This paper presents a new technique for overcoming several types of speech recognition errors by post-processing the output of a continuous speech recognizer. The post-processor output contains fewer errors, thereby making interpretation by higher-level modules, such as a parser, in a speech understanding system more reliable. The primary advantage to the post-processing approach over existing approaches for overcoming SR errors lies in its abilityto introduce options that are not available in the SR module’s output. This work provides evidence for the claim that a modern continuous speech recognizer can be used successfully in “black-box” fashion for robustly interpreting spontaneous utterances …

Adaptive Integration Of Audio And Visual Information Using Discrete And Semi-Continuous Hidden Markov Models In Audiovisual Automatic Speech Recognition, Qin Su

Electrical & Computer Engineering Theses & Dissertations

An audiovisual semi-continuous hidden Markov model (HMM)-based Automatic Speech Recognition (ASR) system and an improved method of integrating audio and visual information in an audiovisual discrete HMM-based ASR system are investigated.

In the audiovisual discrete HMM, an adaptive integration formulation is employed, which incorporates the integration into the HMM at a pre-categorical stage. A visual weighting parameter is determined automatically, which allows the relative contribution of audio and visual information to be adjusted adaptively. Using an adaptive weight, the accuracy increased by 13% compared to the same model with no adaptive weight.

The semi-continuous HMM is a class of models …

Speech Feature Computation For Visual Speech Articulation Training, Stefan Auberg

Electrical & Computer Engineering Theses & Dissertations

A new version of the Visual Speech Articulation Training Aid was implemented which provides real-time visual feedback for speech articulation training. This version was migrated to a Windows Multimedia PC from a previous version which used a Personal Computer (PC) system and additional specialized and expensive custom hardware. This new version will make the system more available to schools and private users. A new method was developed for computing spectral/temporal features which characterize the speech sounds. In addition to the typical frame-based spectral analysis parameters, the new method represents temporal changes of the spectrum with a small number of features. …

Generalized Hidden Filter Markov Models Applied To Speaker Recognition, John M. Colombi

Theses and Dissertations

Classification of time series has wide Air Force, DoD and commercial interest, from automatic target recognition systems on munitions to recognition of speakers in diverse environments. The ability to effectively model the temporal information contained in a sequence is of paramount importance. Toward this goal, this research develops theoretical extensions to a class of stochastic models and demonstrates their effectiveness on the problem of text-independent (language constrained) speaker recognition. Specifically within the hidden Markov model architecture, additional constraints are implemented which better incorporate observation correlations and context, where standard approaches fail. Two methods of modeling correlations are developed, and their …

Unsupervised Algorithms For Learning Emergent Spatio-Temporal Correlations, Chaitanya Tumuluri

Electrical Engineering and Computer Science - Technical Reports

Many applications require the extraction of spatiotemporal correlations among dynamically emergent features of non-stationary distributions. In such applications it is not possible to obtain an a priori analytical characterization of the emergent distribution. This paper extends the Growing Cell Structures (GCS) network and presents two novel (GIST and GEST) networks, which combine unsupervised feature-extraction and Hebbian learning, for tracking such emergent correlations. The networks were successfully tested on the challenging Data Mapping problem, using an execution driven simulation of their implementation in hardware. The results of the simulations show the successful use of the GIST and GEST networks for extracting …

Semantic And Schematic Similarities Between Database Objects: A Context-Based Approach, Vipul Kashyap, Amit P. Sheth

Kno.e.sis Publications

In a multidatabase system, schematic conflicts between two objects are usually of interest only when the objects have some semantic similarity. We use the concept of semantic proximity, which is essentially an abstraction/mapping between the domains of the two objects associated with the context of comparison. An explicit though partial context representation is proposed and the specificity relationship between contexts is defined. The contexts are organized as a meet semi-lattice and associated operations like the greatest lower bound are defined. The context of comparison and the type of abstractions used to relate the two objects form the basis of a …

Negotiation As A Resource Allocation Process, Fernando Tohme

All Computer Science and Engineering Research

The main economic-theoretic approcahes to the problem of resource allocation make little if any reference to negotiation processes. These processes are fundamentally linguistic, based on the exchange of messages among agents. Communication being so fundamental in the characterization of negotiation processes, the analysis of negotiation must emphasize on the structure of the language in which the negotiations take place. Computer science and particularly Artificial Intelligence have provided interesting insights about that linguistic structure. In the first part of this work we present a brief survey of the literature on resource allocation processes in which communication among agents plays a relevant …

An Investigation Into An Effective Method Of Automatically Analysing Oracle Applications To Count Function Points, J. L. Wong

Theses : Honours

Function Point Analysis (FPA) is a synthetic software estimation metric used for computing the size and complexity of applications. It was first introduced by Allan. J. Albrecht during the mid-seventies, as a result of a lengthy research based on applications that were developed using COBOL and PL/1 programming languages. The purpose of this research· is to investigate the possibility, and the most effective method, of automatically performing a Function Point Analysis on Oracle applications that consist of Oracle Forms and Oracle Reports. The research revealed a seemingly lack of other researches on this topic. As FPA was invented a few …

A Systematic Tradeoff Methodology For Acquiring And Validating Imprecise Requirements, J. Yen, Xiaoqing Frank Liu, W. A. Tiao

Computer Science Faculty Research & Creative Works

Requirement analysis is one of the most important phases in a software development process. Existing requirement methodologies are limited in specifying requirements that are usually vague and imprecise, and in supporting tradeoff analysis between the conflicting requirements. In this paper, the elasticity of imprecise requirements is captured using fuzzy logic to facilitate tradeoffs between conflicting requirements. Based on the marginal rate of substitution in decision science, we have developed a systematic approach to elicit the structures and the parameters of imprecise requirements, to validate the scheme for aggregating requirements, and to assess relative priorities of conflicting requirements.

An Adaptive Fuzzy Logic Controller For Intelligent Networking And Control, Irshad Nainar

Theses: Doctorates and Masters

In this thesis, we present a fuzzy logic control scheme to regulate the flow of traffic approaching a set of intersections. An adaptive Fuzzy Logic Traffic Controller (FLTC) is used to adjust the green phase split of the north-south and east-west approaches of a set of traffic signals based on the actual traffic approaching the intersection. Each intersection is coordinated with its neighbouring intersections by adjusting the offset of the local intersection. The offset is adjusted by a local fuzzy logic controller loacted at each intersection. A new fuzzy control scheme, using a supervisory Fuzzy Logic Controller, is also proposed …

On The Impact Of Forgetting On Learning Machines, Rūsiņš Freivalds, Efim Kinber, Carl H. Smith

School of Computer Science & Engineering Faculty Publications

People tend not to have perfect memories when it comes to learning, or to anything else for that matter. Most formal studies of learning, however, assume a perfect memory. Some approaches have restricted the number of items that could be retained. We introduce a complexity theoretic accounting of memory utilization by learning machines. In our new model, memory is measured in bits as a function of the size of the input. There is a hierarchy of learnability based on increasing memory allotment. The lower bound results are proved using an unusual combination of pumping and mutual recursion theorem arguments. For …

Automatic Office Document Classification And Information Extraction, Xiaolong Hao

Dissertations

TEXPR.OS (TEXt PROcessing System) is a document processing system (DPS) to support and assist office workers in their daily work in dealing with information and document management. In this thesis, document classification and information extraction, which are two of the major functional capabilities in TEXPROS, are investigated.

Based on the nature of its content, a document is divided into structured and unstructured (i.e., of free text) parts. The conceptual and content structures are introduced to capture the semantics of the structured and unstructured part of the document respectively. The document is classified and information is extracted based on the analyses …

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 …

Frequency Of Pattern Occurences In A (Dna) Sequence, Mireille Régnier, Wojciech Szpankowski

Department of Computer Science Technical Reports

No abstract provided.

On Pattern Occurrences In A Random Text, Ioannis Fudos, Evaggelia Pitoura, Wojciech Szpankowski

Department of Computer Science Technical Reports

No abstract provided.

A Software Architecture To Support Misuse Intrusion Detection, Sandeep Kumar, Eugene H. Spafford

Department of Computer Science Technical Reports

No abstract provided.

Syntactic Control Of Interference Revisited, Peter W. O'Hearn, A. J. Power, M. Takeyama, R. D. Tennent

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

In "Syntactic Control of Interference" (POPL, 1978), J. C. Reynolds proposes three design principles intended to constrain the scope of imperative state effects in Algol-like languages. The resulting linguistic framework seems to be a very satisfactory way of combining functional and imperative concepts, having the desirable attributes of both purely functional languages (such as pcf) and simple imperative languages (such as the language of while programs). However, Reynolds points out that the "obvious" syntax for interference control has the unfortunate property that fi-reductions do not always preserve typings. Reynolds has subsequently presented a solution to this problem (ICALP, 1989), but …

Parametricity And Local Variables, Peter W. O'Hearn, R. D. Tennent

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

We propose that the phenomenon of local state may be understood in terms of Strachey 's concept of parametric (i.e., uniform) polymorphism. The intuitive basis for our proposal is the following analogy: a non-local procedure is independent of locally-declared variables in the same way that a parametrically polymorphic function is independent of types to which it is instantiated. A connection between parametricity and representational abstraction was first suggested by J. C. Reynolds. Reynolds used logical relations to formalize this connection in languages with type variables and user-defined types. We use relational parametricity to construct a model for an Algol-like language …

Idea Analysis For The Development Of Clinical Trial Strategies, Roberta S. Horowitz

CCAC Theses and Dissertations

Idea Analysis was investigated to determine its ability to organize scientific information and explain the results of specialists' deliberations in designing new clinical trials. Ideas have long been recognized as the engine of creativity. By focusing on the capture of ideas from the scientific literature, idea analysis procedures enable the arrangement of the information into forms consistent with those developed by subject specialists. The most obvious example is the concept structure. Ideas containing a common frequently occurring term/phrase can be depicted as a primary node in the concept network. Related terms will appear as elements associated with that node. Ideas …

Using Fuzzy Set Theory: Exploring How Generalties Can Pinpoint Where To Look For The Answers, Cosmin Radu, Ralph W. Wilkerson

Computer Science Faculty Research & Creative Works

No abstract provided.

Generalized Probabilistic Reasoning And Empirical Studies On Computational Efficiency And Scalability, Eric P. Baenen

Theses and Dissertations

Expert Systems are tools that can be very useful for diagnostic purposes, however current methods of storing and reasoning with knowledge have significant limitations. One set of limitations involves how to store and manipulate uncertain knowledge: much of the knowledge we are dealing with has some degree of uncertainty. These limitations include lack of complete information, not being able to model cyclic information and limitations on the size and complexity of the problems to be solved. If expert systems are ever going to be able to tackle significant real world problems then these deficiencies must be corrected. This paper describes …

Factors In Human-Computer Interface Design (A Pilot Study), Susan Stewart

Theses and Dissertations

The DoD has budgeted over $9.8 billion for 1995 for information technology, yet many government office workers let their existing systems sit idle. This thesis explores why these computers are sitting idle. This researcher's initial hypothesis was that certain features of the human-computer interface can positively or negatively affect efficiency, retention, and satisfaction level of workers. Although some research is being done in this area, interfaces continue to be of poor quality, especially in the DoD, where long procurement cycles, forced purchases, and limited budgets result in out-of-date software. Intuitively most programmers know the human-computer interface impacts on a person's …

Multiple Model Adaptive Estimation And Head Motion Tracking In A Virtual Environment: An Engineering Approach, James E. Russell

Theses and Dissertations

Software engineering tools and techniques were applied to design and implement an application that reduces lag typically present in virtual environment displays. The application was a Multiple Model Adaptive Estimator (MMAE), composed of three Kalman filters, that predicted head orientation one sample period into the future. The environment rendering software used these predictions to generate the environment display. Each of the filters in the MMAE was designed for a different assumed head motion type (benign, moderate, or heavy), which allowed the MMAE to adapt to changes in head movement characteristics. The use of Ada 9X as an implementation language for …

Objects To The Rescue! Or Httpd: The Next Generation Operating System, Andrew P. Black, Jonathan Walpole

Computer Science Faculty Publications and Presentations

This position paper suggests that object-oriented operating systems may provide the means to meet the ever-growing demands of applications. As an example of a successful OOOS, we cite the http daemon. To support the contention that httpd is in fact an operating system, we observe that it implements uniform naming, persistent objects and an invocation meta-protocol, specifies and implements some useful objects, and provides a framework for extensibility.We also believe that the modularity that is characteristic of OO systems should provide a performance benefit rather than a penalty. Our ongoing work in the Synthetix project at OGI is exploring the …

Skin Cancer Diagnosis Using Hierarchal Neural Networks And Fuzzy Logic, H. C. Lee, F. Ercal

Computer Science Technical Reports

Skin cancer, the most common cancer in the United States that affects about 600,000 Americans every year, accounts for 1% of all cancer deaths. Among all, Malignant Melanoma is the most virulent form of skin cancer that is responsible for 75% of all deaths from skin cancer. In 1992, approximately 32,000 people are expected to develop melanoma and about 6,700 will die. However, even malignant melanoma can be treated successfully if detected in the early phase. Therefore, our research goal is to diagnose skin cancer, especially malignant melanoma.

In this study, only the digitized images obtained from color tumor slides …

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 …

Generic Naming In Generative, Constraint- Based Design, Vallis Capoyleas, Xiangping Chen, Christoph M. Hoffmann

Department of Computer Science Technical Reports

No abstract provided.

The Binational English & Spanish Telecommunications Network, Armando A. Arias Jr.

SSGS Faculty Publications and Presentations

BESTNET was established in the early 1980's, as an effort to link universities on both sides of the U.S.- Mexico border through microwave, satellite and cable television technologies. In the late 1980's BESTNET focused primarily on the development of asynchronous computer mediated learning and teaching in an internationally networked virtual environment. For the past six years (1990's) BESTNET has strengthened its binational ties and continued its "high tech" focus through the development of active or vibrant model technology which is assisting in the creation of an on-line binational university setting that is "borderless" (albeit, seamless to the user). Today, this …

A Fuzzy Logic-Based Foundation For Analyzing Imprecise Conflicting Requirements, J. Yen, Xiaoqing Frank Liu

Computer Science Faculty Research & Creative Works

Imprecise requirements are represented by the canonical form in test-score semantics. The concepts of feasibility, satisfiability, and specificity are formalized based on the fuzzy sets. The relationships between requirements are classified to be conflicting and cooperative. A feasible overall requirement can thus be formulated based on the tradeoff analysis of the conflicting requirements by using fuzzy multi-criteria optimization technique

Vic*: A Preprocessor For Virtual-Memory C*, Thomas H. Cormen, Alex Colvin

Computer Science Technical Reports

This paper describes the functionality of ViC*, a compiler-like preprocessor for out-of-core C*. The input to ViC* is a C* program but with certain shapes declared \verb`outofcore`, which means that all parallel variables of these shapes reside on disk. The output is a standard C* program with the appropriate I/O and library calls added for efficient access to out-of-core parallel variables.