Computer Engineering Commons^™

Hierarchical Bayesian Data Fusion Using Autoencoders, Yevgeniy Vladimirovich Reznichenko

Master's Theses (2009 -)

In this thesis, a novel method for tracker fusion is proposed and evaluated for vision-based tracking. This work combines three distinct popular techniques into a recursive Bayesian estimation algorithm. First, semi supervised learning approaches are used to partition data and to train a deep neural network that is capable of capturing normal visual tracking operation and is able to detect anomalous data. We compare various methods by examining their respective receiver operating conditions (ROC) curves, which represent the trade off between specificity and sensitivity for various detection threshold levels. Next, we incorporate the trained neural networks into an existing data …

Search Methods For Mobile Manipulator Performance Measurement, Samuel Nana Yaw Amoako-Frimpong

Master's Theses (2009 -)

Mobile manipulators are a potential solution to the increasing need for additional flexibility and mobility in industrial robotics applications. However, they tend to lack the accuracy and precision achieved by fixed manipulators, especially in scenarios where both the manipulator and the autonomous vehicle move simultaneously. This thesis analyzes the problem of dynamically evaluating the positioning error of mobile manipulators. In particular, it investigates the use of Bayesian methods to predict the position of the end-effector in the presence of uncertainty propagated from the mobile platform. Simulations and real-world experiments are carried out to test the proposed method against a deterministic …

Uncertainty Aware Mapping Of Embedded Systems For Reliability, Performance, And Energy, Wenkai Guan

Master's Theses (2009 -)

Due to technology downscaling, embedded systems have increased in complexity and heterogeneity. The increasingly large process, voltage, and temperature variations negatively affect the design and optimization process of these systems. These factors contribute to increased uncertainties that in turn undermine the accuracy and effectiveness of traditional design approaches. In this thesis, we formulate the problem of uncertainty aware mapping for multicore embedded system platforms as a multi-objective optimization problem. We present a solution to this problem that integrates uncertainty models as a new design methodology constructed with Monte Carlo and evolutionary algorithms. The solution is uncertainty aware because it is …

Development Of Kinematic Templates For Automatic Pronunciation Assessment Using Acoustic-To-Articulatory Inversion, Deriq K. Jones

Master's Theses (2009 -)

Computer-aided pronunciation training (CAPT) is a subcategory of computer-aided language learning (CALL) that deals with the correction of mispronunciation during language learning. For a CAPT system to be effective, it must provide useful and informative feedback that is comprehensive, qualitative, quantitative, and corrective. While the majority of modern systems address the first 3 aspects of feedback, most of these systems do not provide corrective feedback. As part of the National Science Foundation (NSF) funded study “RI: Small: Speaker Independent Acoustic-Articulator Inversion for Pronunciation Assessment”, the Marquette Speech and Swallowing Lab and Marquette Speech and Signal Processing Lab are conducting a …

Deep Neural Networks As Time Series Forecasters Of Energy Demand, Gregory Merkel

Master's Theses (2009 -)

Short-term load forecasting is important for the day-to-day operation of natural gas utilities. Traditionally, short-term load forecasting of natural gas is done using linear regression, autoregressive integrated moving average models, and artificial neural networks. Many purchasing and operating decisions are made using these forecasts, and there can be high cost to both natural gas utilities and their customers if the short-term load forecast is inaccurate. Therefore, the GasDay lab continues to explore new ways to make better forecasts. Recently, deep neural networks (DNNs) have emerged as a powerful tool in machine learning problems. DNNs have been shown to greatly outperform …

Network-On-Chip Based H.264 Video Decoder On A Field Programmable Gate Array, Ian Barge

Master's Theses (2009 -)

This thesis develops the first fully network-on-chip (NoC) based h.264 video decoder implemented in real hardware on a field programmable gate array (FPGA). This thesis starts with an overview of the h.264 video coding standard and an introduction to the NoC communication paradigm. Following this, a series of processing elements (PEs) are developed which implement the component algorithms making up the h.264 video decoder. These PEs, described primarily in VHDL with some Verilog and C, are then mapped to an NoC which is generated using the CONNECT NoC generation tool. To demonstrate the scalability of the proposed NoC based design, …

Data Fusion For Vision-Based Robotic Platform Navigation, Andrés F. Echeverri

Master's Theses (2009 -)

Data fusion has become an active research topic in recent years. Growing computational performance has allowed the use of redundant sensors to measure a single phenomenon. While Bayesian fusion approaches are common in general applications, the computer vision community has largely relegated this approach. Most object following algorithms have gone towards pure machine learning fusion techniques that tend to lack flexibility. Consequently, a more general data fusion scheme is needed. The motivation for this work is to propose methods that allow for the development of simple and cost effective, yet robust visual following robots capable of tracking a general object …

Mims: Mobile Interruption Management System, Mohammad Kayes Kaykobad

Master's Theses (2009 -)

Mobile devices are increasing in an astronomical rate throughout the world. While it is bringing a lot of comfort to the users it is not coming without any hazards. Now a user is susceptible to mobile call interruptions wherever he is, whether he is in the middle of a very important discussion or in a very important task like performing a complicated emergency operation in a hospital. As a result researchers have been studying to find ways to minimize cost of mobile interruptions. In this thesis we have proposed a mobile interruption management system in which callers have been grouped …

Real-Time Topic And Sentiment Analysis In Human-Robot Conversation, Elise Russell

Master's Theses (2009 -)

Socially interactive robots, especially those designed for entertainment and companionship, must be able to hold conversations with users that feel natural and engaging for humans. Two important components of such conversations include adherence to the topic of conversation and inclusion of affective expressions. Most previous approaches have concentrated on topic detection or sentiment analysis alone, and approaches that attempt to address both are limited by domain and by type of reply. This thesis presents a new approach, implemented on a humanoid robot interface, that detects the topic and sentiment of a user’s utterances from text-transcribed speech. It also generates domain-independent, …

Converting Medical Service Provider Data Into A Unified Format For Processing, Brandon Krugman

Master's Theses (2009 -)

Most organizations process flat files regularly. There are different options for processing files, including SQL Server Integration Services (SSIS), BizTalk, SQL import job, and other Extract, Transform, and Load (ETL) processes. All of these options have very strict requirements for file formats. If the format of the file changes, all of these options throw a catastrophic error, and implementing a fix to handle the new format is difficult. With each of the methods, the new format needs to be configured in the development environment, and the data flow must be modified to process all of the changes. Due to the …

Real Time Control Framework Using Android, Aaron Pittenger

Master's Theses (2009 -)

One potential application for a smartphone-type device is a flight management and control computer for an unmanned aerial vehicle (UAV). The hardware employed in most smartphones and tablets has the capabilities necessary to fly an air vehicle without user interaction. The user can pre-program in a flight plan and the smartphone will do the rest. In the past, this real time control application has been done using many separate sensor packages and processors, but never on a single, stand-alone device. Also, capabilities such as the high definition camera present on most smartphones can take photographs and store them on the …

Real-Time Transport Of Internet Telephony Service Utilizing Embedded Resource-Constrained Systems, Kyle Persohn

Master's Theses (2009 -)

No abstract provided.

Propagation Of Plate Acoustic Waves In Contact With Fluid Medium, Nagaraj Ghatadi Suraji

Master's Theses (2009 -)

The characteristics of acoustic waves propagating in thin piezoelectric plates in the presence of a fluid medium contacting one or both of the plate surfaces are investigated. If the velocity of plate wave in the substrate is greater than velocity of bulk wave in the fluid, then a plate acoustic wave (PAW) traveling in the substrate will radiate a bulk acoustic wave (BAW) in the fluid. It is found that, under proper conditions, efficient conversion of energy from plate acoustic waves to bulk acoustic waves and vice versa can be obtained. For example, using the fundamental anti symmetric plate wave …

Design Of A Transtibial Prosthesis Utilizing Active And Passive Components In Conjunction With A Four-Bar Mechanism, Bryan Bergelin

Master's Theses (2009 -)

This thesis outlines the design and testing of a powered ankle prosthesis which utilizes a four-bar mechanism in conjunction with a spiral torsion spring and a servo motor that mimics non-amputee (normal) ankle moments during the stance phase of gait. Previous research has shown that this novel approach is feasible, but the proof-of-concept prototype was not designed to be able to be worn by an amputee due to size, strength and mobility limitations. The goal of this research was to redesign the transtibial prosthesis to fit the requirements of the targeted amputee population. A preliminary virtual prosthesis was designed using …

A Voip Implementation On An Embedded Platform, Zachary D. Lund

Master's Theses (2009 -)

This thesis presents a method of building an open VoIP telephone using off-the-shelf hardware. VoIP has become an important area of study in computer science and engineering, but many of the pieces are expensive and proprietary. We discuss the process of building an open IP telephone, the design decisions and difficulties, a performance evaluation of the different pieces of the system, and methods and suggestions for improving the overall system.

The IP telephone is built on the Freescale 68HC12 microcontroller because of its analog and digital capabilities and the Linksys WRT54GL router for its embedded simplicity and network capabilities. The …

Bi-Planar Postural Stability Model: Fitting Parameters To Patient Data Automatically, Andrew Michael Sovol

Master's Theses (2009 -)

Studies of children with cerebral palsy (CP) have shown that maintaining control of posture in quiet standing is often difficult. Physical therapy and assistive devices, both with or without surgery, may be employed in an effort to improve postural stability. Assessing the effectiveness of such interventions is important for healthcare providers and their patients. A previously created bi-planar model of posture control has been further developed as a method of assessment. This MATLAB and Simulink model has successfully replicated experimental results for typical adults and children, as well as children with CP across three different test conditions: eyes open, eyes …