Computer Engineering Commons^™

Bridging Design And Perception: Novel Tools And Technologies For Creating Effective Human-Robot Interactions, Benjamin Dossett

Electronic Theses and Dissertations

This thesis explores human perception of robots through the use of novel tools and technologies. First, the impact of Augmented Reality (AR) data presentation on human perception of robots is investigated. A study conducted with the AR human-robot teaming system found that robot performance significantly influenced participants’ perceptions, overshadowing the impact of matching or mismatching robot confidence feedback. Second, the DU Want to Build-A-Bot platform is presented, which enables participatory robot design and opens the door for novel research of how robot design affects human perception. The Build-A-Bot platform enables the collection of diverse robot designs, facilitating machine learning analysis …

In-Training Explainability Frameworks To Make Black-Box Machine Learning Models More Explainable., Asuman Cagla Acun

Electronic Theses and Dissertations

Despite ongoing efforts to make black-box machine learning models more explainable, transparent, and trustworthy, there is growing advocacy for using only inherently interpretable models for high-stake decision-making. Post-hoc explanations have recently been criticized for learning surrogate models that may not accurately reflect the actual mechanisms of the original model and for adding computational burden at prediction time. We propose two novel explainability approaches to address these limitations: pre-hoc explainability and co-hoc explainability. These approaches integrate explanations derived from an inherently interpretable white-box model into the learning stage of the black-box model without compromising accuracy. Unlike post-hoc methods, our approach does …

Attention Guided Data Augmentation For Improving The Classification Performance Of Vision Transformers., Nada Baili

Electronic Theses and Dissertations

For over a decade, Deep Neural Networks (DNNs) have been rapidly progressing and achieving great success, forming a robust foundation of state of the art machine learning algorithms that impacted various domains. The advances in data acquisition and processing have undeniably played a major role in these breakthroughs. Data is a crucial component in building successful DNNs, as it enables machine learning models to optimize complex architectures, necessary to perform certain difficult tasks. However, acquiring large-scale data sets is not enough to learn robust models with generalizable features. Instead, an ideal training set should be diverse enough and contain enough …

Multithreaded Applications On The Heterogeneous Research Computing Environment., Sungbo Jung

Electronic Theses and Dissertations

Bioinformatics is a domain that has experienced rapid research growth in recent years, as evidenced by the increasing number of articles in biomedical databases such as PubMed, which adds over a million publications every year. However, this also poses a challenge for researchers who need to find relevant citations for their work. Therefore, developing efficient indexing and searching methods for text data is crucial for Bioinformatics. One key technique for information retrieval is document inversion, which involves creating an inverted index to enable efficient searching through vast collections of text or documents. This Ph.D. research aims to design the research …

Data-Driven Approaches For Enhancing Power Grid Reliability, Behrouz Sohrabi

Electronic Theses and Dissertations

This thesis explores the transformative potential of data-driven approaches in addressing key operational and reliability issues in power systems. The first part of this thesis addresses a prevalent problem in power distribution networks: the accurate identification of load phases. This study develops a data-driven model leveraging consumption measurements from smart meters and corresponding substation data to reconstruct topology information in low-voltage distribution networks. The proposed model is extensively tested using a dataset with more than 5,000 real load profiles, demonstrating satisfactory performance for large-scale networks. The second part of the thesis pivots to a crucial safety concern: the risk and …

Parameter Identifiability Analysis And Calibration For Power System Dynamic Models, Lei Wang

Electronic Theses and Dissertations

Accurate dynamic models are crucial for stable and reliable power system operation, making their validation essential. The accuracy of parameter estimation is influenced by several factors, including available observables, model structure, and estimation methods. However, in dynamic models, estimating parameters can be difficult when some of them are unidentifiable, meaning their values cannot be uniquely determined from the available data. This challenge may stem from limited impact of parameters on the measured outputs, parameter interactions, or poor data quality. Therefore, an identifiability analysis should be conducted prior to parameter estimation or model calibration. This preliminary step helps to determine which …

Designing Ris-Assisted Uav 3d Trajectory Using Deep Reinforcement Learning, Linsong Li

Electronic Theses and Dissertations

Unmanned aerial vehicles (UAVs) are increasingly employed as temporary base stations or access points to facilitate data transfer between ground terminals (GTs). However, in urban environments, UAV-GT communication links often face challenges due to obstructions from buildings and other obstacles, resulting in reduced data transfer efficiency. Reconfigurable intelligent surfaces (RIS) provide a promising solution by reflecting signals to enhance communication quality between UAVs and GTs. This thesis addresses the critical challenge of responsive UAV trajectory optimization in RIS-assisted communication networks. A novel approach is proposed, integrating federated learning with reinforcement learning techniques, specifically Double Deep Q-Network (DDQN) and Deep Deterministic …

Evaluating The Impact Of Perceptual Loss In Generative Adversarial Models And Diffusion Models For Document Image Enhancement, Farzaneh Karimpour

Electronic Theses and Dissertations

Documents often suffer from various types of degradation which make them difficult to read and restrict OCR performance. This study investigates the effectiveness of perceptual loss in enhancing document image cleanup by comparing a GAN-based model and a diffusion model. In our experiments, we utilized the DE-GAN model as a GAN-based model and the NAF-DPM model as a diffusion model, both enhanced by incorporating perceptual loss. We then compared the results of both models and evaluated them by using the DIBCO 2013, DIBCO 2017, and H-DIBCO 2018 datasets revealed that our approach consistently outperforms existing state-of-the-art methods. Results showed that …

Comparative Analysis Of Deep Learning-Based Anomaly Detection Models For Gps Spoofing Detection, Hasan Mirzakhaninafchi

Electronic Theses and Dissertations

As autonomous vehicles (AVs) become integral to modern transportation, their susceptibility to cyber-attacks, particularly GPS spoofing, presents a serious security threat. This study addresses these challenges by applying a suite of deep learning models to enhance the detection of anomalous GPS signals. Focusing on autoencoder-based architectures, the proposed models such as long short-term memory-based variational autoencoder (LSTM-VAE), LSTM-based autoencoder (LSTM-AE), multilayer perceptron-based variational autoencoder (MLP-VAE), MLP-based Autoencoder (MLPAE), Stacked LSTM-based variational autoencoder (Stacked-LSTM-VAE), stacked LSTM-based autoencoder (Stacked-LSTM-AE), memory-augmented-LSTM-VAE (Mem-LSTM-VAE), and time-series-anomaly-detection-generative-adversarial-networks (TadGAN) were trained exclusively on authentic GPS data. This unsupervised learning approach which used for the above-mentioned models enables …

An Integrated Hybrid P2p-Dr Networks For A Transactive Energy Market Platform Considering Electricity Network Constraints, Sheroze Liaquat

Electronic Theses and Dissertations

No abstract provided.

A Modular Framework For Surface-Embedded Actuation And Optical Sensing In Soft Robots., Paul Bupe Jr

Electronic Theses and Dissertations

This dissertation explores the development and integration of modular technologies in soft robotics, with a focus on the OptiGap sensor system. OptiGap serves as a simple, flexible, cost-effective solution for real-time sensing of bending and deformation, validated through simulation and experimentation. Working as part of an emerging category of soft robotics called Soft, Curved, Reconfigurable, Anisotropic Mechanisms, or SCRAMs, this research also introduces the Thermally-Activated SCRAM Limb (TASL) technology, which employs shape-memory alloy (SMA) wire embedded in curved sheets for surface actuation and served as the initial inspiration for OptiGap. In addition, the EneGate system is presented as a complementary …

Convolution And Autoencoders Applied To Nonlinear Differential Equations, Noah Borquaye

Electronic Theses and Dissertations

Autoencoders, a type of artificial neural network, have gained recognition by researchers in various fields, especially machine learning due to their vast applications in data representations from inputs. Recently researchers have explored the possibility to extend the application of autoencoders to solve nonlinear differential equations. Algorithms and methods employed in an autoencoder framework include sparse identification of nonlinear dynamics (SINDy), dynamic mode decomposition (DMD), Koopman operator theory and singular value decomposition (SVD). These approaches use matrix multiplication to represent linear transformation. However, machine learning algorithms often use convolution to represent linear transformations. In our work, we modify these approaches to …

Controllable Language Generation Using Deep Learning, Rohola Zandie

Electronic Theses and Dissertations

The advent of deep neural networks has sparked a revolution in Artificial Intelligence (AI), notably with the creation of Transformer models like GPT-X and ChatGPT. These models have surpassed previous methods in various Natural Language Processing (NLP) tasks. As the NLP field evolves, there is a need to further understand and question the capabilities of these models. Text generation, a crucial part of NLP, remains an area where our comprehension is limited while being critical in research.

This dissertation focuses on the challenging problem of controlling the general behaviors of language models such as sentiment, topical focus, and logical reasoning. …

Terrain And Adversary-Aware Autonomous Robot Navigation, Aniekan Ufot Inyang

Electronic Theses and Dissertations

In autonomous robot navigation, the robot is able to understand the environment around it for intelligent navigation. From its world model of this environment, it generates a global plan for navigation from a position to a goal based on different factors. This research aims to implement autonomous robot navigation by learning terrain affordances: traversability (moving quickly) and concealment (staying hidden from an adversary) using the Preference-based Inverse Reward Learning (PbIRL) methodology. The PbIRL methodology reduces the barrier of generating initial demonstration data to learn the terrain affordances by using a human expert’s preferences to learn individual weights over the terrain …

A Data-Driven Multi-Regime Approach For Predicting Real-Time Energy Consumption Of Industrial Machines., Abdulgani Kahraman

Electronic Theses and Dissertations

This thesis focuses on methods for improving energy consumption prediction performance in complex industrial machines. Working with real-world industrial machines brings several challenges, including data access, algorithmic bias, data privacy, and the interpretation of machine learning algorithms. To effectively manage energy consumption in the industrial sector, it is essential to develop a framework that enhances prediction performance, reduces energy costs, and mitigates air pollution in heavy industrial machine operations. This study aims to assist managers in making informed decisions and driving the transition towards green manufacturing. The energy consumption of industrial machinery is substantial, and the recent increase in CO2 …

Blockchain Security: Double-Spending Attack And Prevention, William Henry Scott Iii

Electronic Theses and Dissertations

This thesis shows that distributed consensus systems based on proof of work are vulnerable to hashrate-based double-spending attacks due to abuse of majority rule. Through building a private fork of Litecoin and executing a double-spending attack this thesis examines the mechanics and principles behind the attack. This thesis also conducts a survey of preventative measures used to deter double-spending attacks, concluding that a decentralized peer-to-peer network using proof of work is best protected by the addition of an observer system whether internal or external.

Dynamic Scene Understanding: Pedestrian Tracking From Aerial Devices., Abdelhamid Bouzid

Electronic Theses and Dissertations

Multiple Object Tracking (MOT) is the problem that involves following the trajectory of multiple objects in a sequence, generally a video. Pedestrians are among the most interesting subjects to track and recognize for many purposes such as surveillance, and safety. In the recent years, Unmanned Aerial Vehicles (UAV’s) have been viewed as a viable option for monitoring public areas, as they provide a low-cost method of data collection while covering large and difficult-to-reach areas. In this thesis, we present an online pedestrian tracking and re-identification from aerial devices framework. This framework is based on learning a compact directional statistic distribution …

Guided Data Augmentation For Improved Semi-Supervised Image Classification In Low Data Regime., Fadoua Khmaissia

Electronic Theses and Dissertations

Deep learning models have achieved state of the art performances, especially for computer vision applications. Much of the recent successes can be attributed to the existence of large, high quality, labeled datasets. However, in many real-world applications, collecting similar datasets is often cumbersome and time consuming. For instance, developing robust automatic target recognition models from infrared images still faces major challenges. This is mainly due to the difficulty of acquiring high resolution inputs, sensitivity to the thermal sensors' calibration, meteorological conditions, targets' scale and viewpoint invariance. Ideally, a good training set should contain enough variations within each class for the …

Modeling, Simulation And Control Of Microrobots For The Microfactory., Zhong Yang

Electronic Theses and Dissertations

Future assembly technologies will involve higher levels of automation in order to satisfy increased microscale or nanoscale precision requirements. Traditionally, assembly using a top-down robotic approach has been well-studied and applied to the microelectronics and MEMS industries, but less so in nanotechnology. With the boom of nanotechnology since the 1990s, newly designed products with new materials, coatings, and nanoparticles are gradually entering everyone’s lives, while the industry has grown into a billion-dollar volume worldwide. Traditionally, nanotechnology products are assembled using bottom-up methods, such as self-assembly, rather than top-down robotic assembly. This is due to considerations of volume handling of large …

Metabolomic Differentiation Of Tumor Core And Edge In Glioma., Mary E. Baxter

Electronic Theses and Dissertations

Glioma is one of the most aggressive forms of brain cancer. It has been shown that the microenvironments differ significantly between the core and edge regions of glioma tumors. This study obtained metabolomic profiles of glioma core and edge regions using paired glioma core and edge tissue samples from 27 human patients. Data was acquired by performing liquid-liquid metabolite extraction and 2DLC-MS/MS on the tissue samples. In addition, a boosted generalized linear machine learning model was employed to predict the metabolomic profiles associated with O-6-methylguanine-DNA methyltransferase (MGMT) promoter methylation.

A panel of 66 metabolites was found to be statistically significant …

Reference Frames In Human Sensory, Motor, And Cognitive Processing, Dongcheng He

Electronic Theses and Dissertations

Reference-frames, or coordinate systems, are used to express properties and relationships of objects in the environment. While the use of reference-frames is well understood in physical sciences, how the brain uses reference-frames remains a fundamental question. The goal of this dissertation is to reach a better understanding of reference-frames in human perceptual, motor, and cognitive processing. In the first project, we study reference-frames in perception and develop a model to explain the transition from egocentric (based on the observer) to exocentric (based outside the observer) reference-frames to account for the perception of relative motion. In a second project, we focus …

Design, Determination, And Evaluation Of Gender-Based Bias Mitigation Techniques For Music Recommender Systems, Sunny Shrestha

Electronic Theses and Dissertations

The majority of smartphone users engage with a recommender system on a daily basis. Many rely on these recommendations to make their next purchase, download the next game, listen to the new music or find the next healthcare provider. Although there are plenty of evidence backed research that demonstrates presence of gender bias in Machine Learning (ML) models like recommender systems, the issue is viewed as a frivolous cause that doesn’t merit much action. However, gender bias poses to effect more than half of the population as by default ML systems are designed to cater to a cisgender man. This …

Artificial Emotional Intelligence In Socially Assistive Robots, Hojjat Abdollahi

Electronic Theses and Dissertations

Artificial Emotional Intelligence (AEI) bridges the gap between humans and machines by demonstrating empathy and affection towards each other. This is achieved by evaluating the emotional state of human users, adapting the machine’s behavior to them, and hence giving an appropriate response to those emotions. AEI is part of a larger field of studies called Affective Computing. Affective computing is the integration of artificial intelligence, psychology, robotics, biometrics, and many more fields of study. The main component in AEI and affective computing is emotion, and how we can utilize emotion to create a more natural and productive relationship between humans …

Iot In Smart Communities, Technologies And Applications., Muhammad Zaigham Abbas Shah Syed

Electronic Theses and Dissertations

Internet of Things is a system that integrates different devices and technologies, removing the necessity of human intervention. This enables the capacity of having smart (or smarter) cities around the world. By hosting different technologies and allowing interactions between them, the internet of things has spearheaded the development of smart city systems for sustainable living, increased comfort and productivity for citizens. The Internet of Things (IoT) for Smart Cities has many different domains and draws upon various underlying systems for its operation, in this work, we provide a holistic coverage of the Internet of Things in Smart Cities by discussing …

Computer Aided Diagnosis System For Breast Cancer Using Deep Learning., Asma Baccouche

Electronic Theses and Dissertations

The recent rise of big data technology surrounding the electronic systems and developed toolkits gave birth to new promises for Artificial Intelligence (AI). With the continuous use of data-centric systems and machines in our lives, such as social media, surveys, emails, reports, etc., there is no doubt that data has gained the center of attention by scientists and motivated them to provide more decision-making and operational support systems across multiple domains. With the recent breakthroughs in artificial intelligence, the use of machine learning and deep learning models have achieved remarkable advances in computer vision, ecommerce, cybersecurity, and healthcare. Particularly, numerous …

Role Of Machine Learning In Early Diagnosis Of Kidney Diseases., Mohamed Nazih Mohamed Ibrahim Shehata

Electronic Theses and Dissertations

Machine learning (ML) and deep learning (DL) approaches have been used as indispensable tools in modern artificial intelligence-based computer-aided diagnostic (AIbased CAD) systems that can provide non-invasive, early, and accurate diagnosis of a given medical condition. These AI-based CAD systems have proven themselves to be reproducible and have the generalization ability to diagnose new unseen cases with several diseases and medical conditions in different organs (e.g., kidneys, prostate, brain, liver, lung, breast, and bladder). In this dissertation, we will focus on the role of such AI-based CAD systems in early diagnosis of two kidney diseases, namely: acute rejection (AR) post …

Reduced Fuel Emissions Through Connected Vehicles And Truck Platooning, Paul D. Brummitt

Electronic Theses and Dissertations

Vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication enable the sharing, in real time, of vehicular locations and speeds with other vehicles, traffic signals, and traffic control centers. This shared information can help traffic to better traverse intersections, road segments, and congested neighborhoods, thereby reducing travel times, increasing driver safety, generating data for traffic planning, and reducing vehicular pollution. This study, which focuses on vehicular pollution, used an analysis of data from NREL, BTS, and the EPA to determine that the widespread use of V2V-based truck platooning—the convoying of trucks in close proximity to one another so as to reduce air drag …

A Water-Surface Self-Leveling Landing Platform For Small-Scale Uavs, Mbidi Santos

Electronic Theses and Dissertations

Because many of the most widely used UAVs, such as the Vertical Take-Off and Landing (VTOL), cannot land securely on sloped or fast-changing surfaces, there is a need to design better deployment and landing stations. This document proposes an approach to design a water-surface self-leveling landing platform by implementing the best concept to be used as a safe ground for UAVs to land and deploy on open waters. After conceptualizing multiple design ideas, these options were laid out in a decision matrix with four criteria: degrees of freedom, mechanical complexity, manufacturing, and cost. The chosen concept was the spherical parallel …

Classification Of Electropherograms Using Machine Learning For Parkinson’S Disease, Soroush Dehghan

Electronic Theses and Dissertations

Parkinson’s disease (PD) is a neurodegenerative movement disorder that progresses gradually over time. The onset of symptoms in people who are suffering from PD can vary from case to case, and it depends on the progression of the disease in each patient. The PD symptoms gradually develop and exacerbate the patient’s movements throughout time. An early diagnosis of PD could improve the outcomes of treatments and could potentially delay the progression of this disorder and that makes discovering a new diagnostic method valuable. In this study, I investigate the feasibility of using a machine learning (ML) approach to classify PD …

Study On Performance Of Pruned Cnn-Based Classification Models, Mengling Deng

Electronic Theses and Dissertations

Convolutional Neural Network (CNN) is a neural network developed for processing image data. CNNs have been studied extensively and have been used in numerous computer vision tasks such as image classification and segmentation, object detection and recognition, etc. [1] Although, the CNNs-based approaches showed humanlevel performances in these tasks [2], they require heavy computation in both training and inference stages, and the models consist of millions of parameters. This hinders the development and deployment of CNN-based models for real world applications. Neural Network Pruning and Compression techniques have been proposed [3, 4] to reduce the computation complexity of trained CNNs …