Engineering Commons^™

Facial Expression Recognition Using Convolutional Neural Networks (Cnns) And Generative Adversarial Networks (Gans) For Data Augmentation And Image Generation, Shekhar Singh

UNLV Theses, Dissertations, Professional Papers, and Capstones

Facial expressions play a crucial role in human communication, serving as a powerful means to convey emotions. However, classifying facial expressions using artificial intelligence (AI) can be challenging, especially with small datasets and images. Facial Expression Recognition (FER) is an active area of research, with Convolutional Neural Networks (CNNs) being widely employed for classification. In this research, we propose a CNN-based approach for FER that utilizes both original and augmented datasets to enhance classification accuracy. Experimental results on the FER2013 dataset show test accuracies of 63.39% and 64.59% for the original and augmented datasets, respectively, in a seven-class classification task. …

Enhanced Iot-Based Electrocardiogram Monitoring System With Deep Learning, Jian Ni

UNLV Theses, Dissertations, Professional Papers, and Capstones

Due to the rapid development of computing and sensing technologies, Internet of Things (IoT)-based cardiac monitoring plays a crucial role in providing patients with cost-efficient solutions for long-term, continuous, and pervasive electrocardiogram (ECG) monitoring outside a hospital setting. In a typical IoT-based ECG monitoring system, ECG signals are picked up by sensors located on the edge, and then uploaded to the remote cloud servers. ECG interpretation is performed for the collected ECGs in the cloud servers and the analysis results can be made instantly available to the patients as well as their healthcare providers.In this dissertation, we first examine the …

Information-Theoretic Model Diagnostics (Infomod), Armin Esmaeilzadeh

UNLV Theses, Dissertations, Professional Papers, and Capstones

Model validation is a critical step in the development, deployment, and governance of machine learning models. During the validation process, the predictive power of a model is measured on unseen datasets with a variety of metrics such as Accuracy and F1-Scores for classification tasks. Although the most used metrics are easy to implement and understand, they are aggregate measures over all the segments of heterogeneous datasets, and therefore, they do not identify the performance variation of a model among different data segments. The lack of insight into how the model performs over segments of unseen datasets has raised significant challenges …

Identification Of Factors Contributing To Traffic Crashes By Analysis Of Text Narratives, Cristian D. Arteaga-Sanchez

UNLV Theses, Dissertations, Professional Papers, and Capstones

The fatalities, injuries, and property damage that result from traffic crashes impose a significant burden on society. Current research and practice in traffic safety rely on analysis of quantitative data from crash reports to understand crash severity contributors and develop countermeasures. Despite advances from this effort, quantitative crash data suffers from drawbacks, such as the limited ability to capture all the information relevant to the crashes and the potential errors introduced during data collection. Crash narratives can help address these limitations, as they contain detailed descriptions of the context and sequence of events of the crash. However, the unstructured nature …

Hyperspectral Image Analysis Of Food For Nutritional Intake, Shirin Nasr Esfahani

UNLV Theses, Dissertations, Professional Papers, and Capstones

The primary object of this dissertation is to investigate the application of hyperspectral technology to accommodate for the growing demand in the automatic dietary assessment applications. Food intake is one of the main factors that contribute to human health. In other words, it is necessary to get information about the amount of nutrition and vitamins that a human body requires through a daily diet. Manual dietary assessments are time-consuming and are also not precise enough, especially when the information is used for the care and treatment of hospitalized patients. Moreover, the data must be analyzed by nutritional experts. Therefore, researchers …

Radiomic Features To Predict Overall Survival Time For Patients With Glioblastoma Brain Tumors Based On Machine Learning And Deep Learning Methods, Lina Chato

UNLV Theses, Dissertations, Professional Papers, and Capstones

Machine Learning (ML) methods including Deep Learning (DL) Methods have been employed in the medical field to improve diagnosis process and patient’s prognosis outcomes. Glioblastoma multiforme is an extremely aggressive Glioma brain tumor that has a poor survival rate. Understanding the behavior of the Glioblastoma brain tumor is still uncertain and some factors are still unrecognized. In fact, the tumor behavior is important to decide a proper treatment plan and to improve a patient’s health. The aim of this dissertation is to develop a Computer-Aided-Diagnosis system (CADiag) based on ML/DL methods to automatically estimate the Overall Survival Time (OST) for …

A Human-Embodied Drone For Dexterous Aerial Manipulation, Dongbin Kim

UNLV Theses, Dissertations, Professional Papers, and Capstones

Current drones perform a wide variety of tasks in surveillance, photography, agriculture, package delivery, etc. However, these tasks are performed passively without the use of human interaction. Aerial manipulation shifts this paradigm and implements drones with robotic arms that allow interaction with the environment rather than simply sensing it. For example, in construction, aerial manipulation in conjunction with human interaction could allow operators to perform several tasks, such as hosing decks, drill into surfaces, and sealing cracks via a drone. This integration with drones will henceforth be known as dexterous aerial manipulation.

Our recent work integrated the worker’s experience into …

Material Handling With Embodied Loco-Manipulation, Jean Chagas Vaz

UNLV Theses, Dissertations, Professional Papers, and Capstones

Material handling is an intrinsic component of disaster response. Typically, first responders, such as firefighters and/or paramedics, must carry, push, pull, and handle objects, facilitating the transportation of goods. For many years, researchers from around the globe have sought to enable full-sized humanoid robots to perform such essential material handling tasks. This work aims to tackle current limitations of humanoids in the realm of interaction with common objects such as carts, wheelbarrows, etc. Throughout this research, many methods will be applied to ensure a stable Zero Moment Point (ZMP) trajectory to allow a robust gait while loco-manipulating a cart. The …

Rotten With Prediction, Serena Raquel Hicks

UNLV Theses, Dissertations, Professional Papers, and Capstones

This project focuses on the relationship between religion and technology as it is portrayed in Science Fiction (SF). This thesis explores the SF genre rhetorically by examining the 2002 movie Minority Report (MR), which signaled the importance of surveillance and the need to predict future crimes following 9/11. The events of 9/11 played a significant role in post 9/11 SF films, which reflect and critique our communal and cultural values. 9/11 created a new relationship between the U.S justice system, predictive technologies (PTs), and data gathering. Through the Bush Doctrine of “preemptive action,” the U.S government attempted to use Dataism, …

Forecasting Pedestrian Trajectory Using Deep Learning, Arsal Syed

UNLV Theses, Dissertations, Professional Papers, and Capstones

In this dissertation we develop different methods for forecasting pedestrian trajectories. Complete understanding of pedestrian motion is essential for autonomous agents and social robots to make realistic and safe decisions. Current trajectory prediction methods rely on incorporating historic motion, scene features and social interaction to model pedestrian behaviors. Our focus is to accurately understand scene semantics to better forecast trajectories. In order to do so, we leverage semantic segmentation to encode static scene features such as walkable paths, entry/exits, static obstacles etc. We further evaluate the effectiveness of using semantic maps on different datasets and compare its performance with already …

A Hyperelastic Porous Media Framework For Ionic Polymer-Metal Composites And Characterization Of Transduction Phenomena Via Dimensional Analysis And Nonlinear Regression, Zakai J. Olsen

UNLV Theses, Dissertations, Professional Papers, and Capstones

Ionic polymer-metal composites (IPMC) are smart materials that exhibit large deformation in response to small applied voltages, and conversely generate detectable electrical signals in response to mechanical deformations. The study of IPMC materials is a rich field of research, and an interesting intersection of material science, electrochemistry, continuum mechanics, and thermodynamics. Due to their electromechanical and mechanoelectrical transduction capabilities, IPMCs find many applications in robotics, soft robotics, artificial muscles, and biomimetics. This study aims to investigate the dominating physical phenomena that underly the actuation and sensing behavior of IPMC materials. This analysis is made possible by developing a new, hyperelastic …

An Investigation Into Multi-View Error Correcting Output Code Classifiers Applied To Organ Tissue Classification, Daniel Alvarez

UNLV Theses, Dissertations, Professional Papers, and Capstones

Large amounts of data is being generated constantly each day, so much data that it is difficult to find patterns in order to predict outcomes and make decisions for both humans and machines alike. It would be useful if this data could be simplified using machine learning techniques. For example, biological cell identity is dependent on many factors tied to genetic processes. Such factors include proteins, gene transcription, and gene methylation. Each of these factors are highly complex mechanism with immense amounts of data. Simplifying these can then be helpful in finding patterns in them. Error-Correcting Output Codes (ECOC) does …

A Framework For Vector-Weighted Deep Neural Networks, Carter Chiu

UNLV Theses, Dissertations, Professional Papers, and Capstones

The vast majority of advances in deep neural network research operate on the basis of a real-valued weight space. Recent work in alternative spaces have challenged and complemented this idea; for instance, the use of complex- or binary-valued weights have yielded promising and fascinating results. We propose a framework for a novel weight space consisting of vector values which we christen VectorNet. We first develop the theoretical foundations of our proposed approach, including formalizing the requisite theory for forward and backpropagating values in a vector-weighted layer. We also introduce the concept of expansion and aggregation functions for conversion between real …

Classification Of Vegetation In Aerial Imagery Via Neural Network, Gevand Balayan

UNLV Theses, Dissertations, Professional Papers, and Capstones

This thesis focuses on the task of trying to find a Neural Network that is best suited for identifying vegetation from aerial imagery. The goal is to find a way to quickly classify items in an image as highly likely to be vegetation(trees, grass, bushes and shrubs) and then interpolate that data and use it to mark sections of an image as vegetation. This has practical applications as well. The main motivation of this work came from the effort that our town takes in conserving water. By creating an AI that can easily recognize plants, we can better monitor the …

Analysis Of Bitcoin Cryptocurrency And Its Mining Techniques, Suman Ghimire

UNLV Theses, Dissertations, Professional Papers, and Capstones

Bitcoin is a peer-to-peer digital, decentralized cryptocurrency created by an individual under pseudonym Satoshi Nakamoto. In fact, it is the first digital, decentralized currency. Several developers and organizations have explored the importance of digital cryptocurrency and the concept of the blockchain. Bitcoin is assumed to be one of the secure and comfortable payment methods that can be used in the upcoming days. The backbone of Bitcoin mining is the concept of the blockchain, which is assumed to beone of the ingenious invention of this century. The blockchain is the collection of blocks that are linked together in such a way …

Multi-Resolution Spatio-Temporal Change Analyses Of Hydro-Climatological Variables In Association With Large-Scale Oceanic-Atmospheric Climate Signals, Kazi Ali Tamaddun

UNLV Theses, Dissertations, Professional Papers, and Capstones

The primary objective of the work presented in this dissertation was to evaluate the change patterns, i.e., a gradual change known as the trend, and an abrupt change known as the shift, of multiple hydro-climatological variables, namely, streamflow, snow water equivalent (SWE), temperature, precipitation, and potential evapotranspiration (PET), in association with the large-scale oceanic-atmospheric climate signals. Moreover, both observed datasets and modeled simulations were used to evaluate such change patterns to assess the efficacy of the modeled datasets in emulating the observed trends and shifts under the influence of uncertainties and inconsistencies. A secondary objective of this study was to …

The Affective Perceptual Model: Enhancing Communication Quality For Persons With Pimd, Jadin Tredup

UNLV Theses, Dissertations, Professional Papers, and Capstones

Methods for prolonged compassionate care for persons with Profound Intellectual and Multiple Disabilities (PIMD) require a rotating cast of import people in the subjects life in order to facilitate interaction with the external environment. As subjects continue to age, dependency on these people increases with complexity of communications while the quality of communication decreases. It is theorized that a machine learning (ML) system could replicate the attuning process and replace these people to promote independence. This thesis extends this idea to develop a conceptual and formal model and system prototype.

The main contributions of this thesis are: (1) proposal of …

Design On High Performance Nanoscale Cmos Circuits With Low Temperature Sensitivity, Ming Zhu

UNLV Theses, Dissertations, Professional Papers, and Capstones

With the rapid development of integrated circuit (IC) design and manufacturing technology, the transistor size now can be shrunk into only couple of nanometers whereas billions of transistors can be squeezed into a square millimeter, providing unprecedented computation power. However, accompanied with continuous device miniaturization and increased integration density is the explosive growth of on-chip power dissipation and a wide range of temperature fluctuation, which can heavily and negatively affect the delay performance of the circuit, or in the worst case, the circuit may malfunction and the system can be unreliable. Therefore, improved performance resilience against temperature variations has become …

Improving Pattern Recognition And Neural Network Algorithms With Applications To Solar Panel Energy Optimization, Ernesto Zamora Ramos

UNLV Theses, Dissertations, Professional Papers, and Capstones

Artificial Intelligence is a big part of automation and with today's technological advances, artificial intelligence has taken great strides towards positioning itself as the technology of the future to control, enhance and perfect automation. Computer vision includes pattern recognition and classification and machine learning. Computer vision is at the core of decision making and it is a vast and fruitful branch of artificial intelligence. In this work, we expose novel algorithms and techniques built upon existing technologies to improve pattern recognition and neural network training, initially motivated by a multidisciplinary effort to build a robot that helps maintain and optimize …

Deployment Strategies For A Fleet Of Unmanned Aerial Vehicles Providing Cellular And Data Services, Nima Mohseni

UNLV Theses, Dissertations, Professional Papers, and Capstones

Wireless voice and data communications have become an essential part of our day to day lives. In order to provide these services to as many people as possible, a great infrastructure has been put in place over the last two decades throughout the world. The current infrastructure is mainly consists of cellular towers with gateways to the telecommunication backbone. The wireless infrastructure is doing an adequate job of providing voice and data services, getting more powerful and efficient every day. However, because wireless infrastructure is mainly based on fixed cell towers, it lacks the flexibility and dynamism that may be …

Vulnerability Analysis And Security Framework For Zigbee Communication In Iot, Charbel Azzi

UNLV Theses, Dissertations, Professional Papers, and Capstones

Securing IoT (Internet of Things) systems in general, regardless of the communication technology used, has been the concern of many researchers and private companies. As for ZigBee security concerns, much research and many experiments have been conducted to better predict the nature of potential security threats. In this research we are addressing several ZigBee vulnerabilities by performing first hand experiments and attack simulations on ZigBee protocol. This will allow us to better understand the security issues surveyed and find ways to mitigate them. Based on the attack simulations performed and the survey conducted, we have developed a ZigBee IoT framework …

Acoustic Detection, Source Separation, And Classification Algorithms For Unmanned Aerial Vehicles In Wildlife Monitoring And Poaching, Carlo Lopez-Tello

UNLV Theses, Dissertations, Professional Papers, and Capstones

This work focuses on the problem of acoustic detection, source separation, and classification under noisy conditions. The goal of this work is to develop a system that is able to detect poachers and animals in the wild by using microphones mounted on unmanned aerial vehicles (UAVs). The classes of signals used to detect wildlife and poachers include: mammals, birds, vehicles and firearms. The noise signals under consideration include: colored noises, UAV propeller and wind noises.

The system consists of three sub-systems: source separation (SS), signal detection, and signal classification. Non-negative Matrix Factorization (NMF) is used for source separation, and random …

Efficient Algorithms For Clustering Polygonal Obstacles, Sabbir Kumar Manandhar

UNLV Theses, Dissertations, Professional Papers, and Capstones

Clustering a set of points in Euclidean space is a well-known problem having applications in pattern recognition, document image analysis, big-data analytics, and robotics. While there are a lot of research publications for clustering point objects, only a few articles have been reported for clustering a given distribution of obstacles. In this thesis we examine the development of efficient algorithms for clustering a given set of convex obstacles in the 2D plane. One of the methods presented in this work uses a Voronoi diagram to extract obstacle clusters. We also consider the implementation issues of point/obstacle clustering algorithms.

A Study On Automated Process For Extracting White Blood Cellular Data From Microscopic Digital Injured Skeletal Muscle Images, Bibek Karki

UNLV Theses, Dissertations, Professional Papers, and Capstones

Skeletal muscle injury is one of the common injuries caused by high-intensity sports activities, military related works, and natural disasters. In order to discover better therapies, it is important to study muscle regeneration process. Muscle regeneration process tracking is the act of monitoring injured tissue section over time, noting white blood cell behavior and cell-fiber relations. A large number of microscopic images are taken for tracking muscle regeneration process over multiple time instances. Currently, manual approach is widely used to analyze a microscopic image of muscle cross section, which is time consuming, tedious and buggy.

Automation of this research methodology …

Object Detection And Tracking In Wide Area Surveillance Using Thermal Imagery, Santosh Bhusal

UNLV Theses, Dissertations, Professional Papers, and Capstones

The main objective behind this thesis is to examine how existing vision-based detection and tracking algorithms perform in thermal imagery-based video surveillance. While color-based surveillance has been extensively studied, these techniques can not be used during low illumination, at night, or with lighting changes and shadows which limits their applicability. The main contributions in this thesis are (1) the creation of a new color-thermal dataset, (2) a detailed performance comparison of different color-based detection and tracking algorithms on thermal data and (3) the proposal of an adaptive neural network for false detection rejection.

Since there are not many publicly available …

Simulation, And Overload And Stability Analysis Of Continuous Time Sigma Delta Modulator, Kyung Kang

UNLV Theses, Dissertations, Professional Papers, and Capstones

The ever increasing demand for faster and more powerful digital applications requires high speed, high resolution ADCs. Currently, sigma delta modulators ADCs are extensively used in broadband telecommunication systems because they are an effective solution for high data-rate wireless communication systems that require low power consumption, high speed, high resolution, and large signal bandwidths.

Because mixed-signal integrated circuits such as Continuous Time sigma delta modulators contain both analog and digital circuits, mixed signal circuits are not as simple to model and simulate as all discrete or all analog systems. In this dissertation, the delta transform is used to simulate CT …

Design And Implementation Of An Instruction Set Architecture And An Instruction Execution Unit For The Rez9 Coprocessor System, Daniel Spencer Anderson

UNLV Theses, Dissertations, Professional Papers, and Capstones

While the use of RNS has provided groundbreaking theory and progress in this field, the applications still lack viable testing platforms to test and verify the theory. This Thesis outlines the processing of developing an instruction set architecture (ISA) and an instruction execution unit (IEU) to help make the first residue based general processor a viable testing platform to address the mentioned problems.

Consider a 32-bit ripple adder. The delay on this device will be 32N where N is the delay for each adder to complete its operation. The delay of this process is due to the need to propagate …

Designing A Biomimetic Testing Platform For Actuators In A Series-Elastic Co-Contraction System, Ryan Tyler Schroeder

UNLV Theses, Dissertations, Professional Papers, and Capstones

Actuators determine the performance of robotic systems at the most intimate of levels. As a result, much work has been done to assess the performance of different actuator systems. However, biomimetics has not previously been utilized as a pretext for tuning a series elastic actuator system with the purpose of designing an empirical testing platform. Thus, an artificial muscle tendon system has been developed in order to assess the performance of two distinct actuator types: (1) direct current electromagnetic motors and (2) ultrasonic rotary piezoelectric motors. Because the design of the system takes advantage of biomimetic operating principles such as …

Modeling And Development Of Human Interface For Pedestrian Simulator, Romesh Khaddar

UNLV Theses, Dissertations, Professional Papers, and Capstones

According to Traveler opinion and perception survey of 2005, 107.4 million Americans use walking as regular mode of travel, which amounts to 51% of American population. In 2009, 4092 pedestrian fatalities have been reported nationwide with a fatality rate of 1.33 which totals 59,000 crashes. Also, pedestrians are over represented in crash data by accounting more than 12% of fatalities but on 10.9% of trips. This makes a perfect case for understanding the causes behind such statistics, calling for a continuous research on pedestrians walking behavior and their interactions with surroundings.

Current research in pedestrian simulation focuses on surveys and …

Gate Monte Carlo Simulations In A Cloud Computing Environment, Blake Austin Rowedder

UNLV Theses, Dissertations, Professional Papers, and Capstones

The GEANT4-based GATE is a unique and powerful Monte Carlo (MC) platform, which provides a single code library allowing the simulation of specific medical physics applications, e.g. PET, SPECT, CT, radiotherapy, and hadron therapy. However, this rigorous yet flexible platform is used only sparingly in the clinic due to its lengthy calculation time. By accessing the powerful computational resources of a cloud computing environment, GATE's runtime can be significantly reduced to clinically feasible levels without the sizable investment of a local high performance cluster. This study investigated a reliable and efficient execution of GATE MC simulations using a commercial cloud …