Computer Engineering Commons^™

Machine Learning Assisted Discovery Of Shape Memory Polymers And Their Thermomechanical Modeling, Cheng Yan

LSU Doctoral Dissertations

As a new class of smart materials, shape memory polymer (SMP) is gaining great attention in both academia and industry. One challenge is that the chemical space is huge, while the human intelligence is limited, so that discovery of new SMPs becomes more and more difficult. In this dissertation, by adopting a series of machine learning (ML) methods, two frameworks are established for discovering new thermoset shape memory polymers (TSMPs). Specifically, one of them is performed by a combination of four methods, i.e., the most recently proposed linear notation BigSMILES, supplementing existing dataset by reasonable approximation, a mixed dimension (1D …

A Component-Based Analysis For Online Proctoring, Salma Roshdy Ali

Theses and Dissertations

The switch to online learning due to the COVID-19 revealed flaws in the existing learning methods, especially with online proctored assessments. Hence, online proctoring using computers was needed for a fair evaluation. Many studies develop cheating detection systems using several approaches. However, to the best of our knowledge, none of the existing studies investigated the impact of their system components in detecting cheating behaviors. Combining system components, even if they do not significantly improve the system performance in cheating detection, can cause an overload on the system. Therefore, our goal is to investigate the system components’ impact, individually and combined, …

Image Provenance Analysis, Daniel Moreira, William Theisen, Walter Scheirer, Aparna Bharati, Joel Brogan, Anderson Rocha

Computer Science: Faculty Publications and Other Works

The literature of multimedia forensics is mainly dedicated to the analysis of single assets (such as sole image or video files), aiming at individually assessing their authenticity. Different from this, image provenance analysis is devoted to the joint examination of multiple assets, intending to ascertain their history of edits, by evaluating pairwise relationships. Each relationship, thus, expresses the probability of one asset giving rise to the other, through either global or local operations, such as data compression, resizing, color-space modifications, content blurring, and content splicing. The principled combination of these relationships unveils the provenance of the assets, also constituting an …

Using Deep Neural Network And Transformers To Extract Graphene Compounds And Properties, Ayman Ibn Jaman

Computer Science Graduate Research Workshop

No abstract provided.

Visual Homing For Robot Teams: Do You See What I See?, Damian Lyons, Noah Petzinger

Faculty Publications

Visual homing is a lightweight approach to visual navigation which does not require GPS. It is very attractive for robot platforms with a low computational capacity. However, a limitation is that the stored home location must be initially within the field of view of the robot. Motivated by the increasing ubiquity of camera information we propose to address this line-of-sight limitation by leveraging camera information from other robots and fixed cameras. To home to a location that is not initially within view, a robot must be able to identify a common visual landmark with another robot that can be used …

Autonomous And Interactive Control Of A Mobile Robot, Dylan Hoover, Tanner Kaczmarek, Kevin Molumphy, Stephen Tambussi

Computer Science and Engineering Senior Theses

An autonomous and interactive control of a mobile robot is a desired asset to the Robotics Systems Laboratory (RSL) at Santa Clara University and to the food automation company, L2F. This benefits their future endeavors of having helpful cobots navigate and assist entities in their respective environments. This paper documents the development of software for a cobot that satisfies the basic requirements for easy and safe control of an autonomous robot within a dynamic environment. The completed software includes the ability for the cobot to track a person and then autonomously follow that person around at a safe following distance …

Operating Machine Learning To Identify Tools (Omlit), Jason Chavez, Grant Schorr, Sebastian De La Cruz

Computer Science and Engineering Senior Theses

Today’s society is heavily reliant on using data to improve systems and create innovative technology. This project takes advantage of artificial intelligence, specifically machine learning (ML) which has allowed us to create a web application that detects tools, specifically hand tools. By having the user upload an image of one of three hand tools (screw driver, handsaw, or power drill) the user will be able to identify the tool as well as be provided information on hand tool safety. This identification system is handled by a pre-trained Convolutional Neural Network (CNN) model and trained using a self built data set. …

Visual Homing For Robot Teams: Do You See What I See?, Damian Lyons, Noah Petzinger

Faculty Publications

Visual homing is a lightweight approach to visual navigation which does not require GPS. It is very attractive for robot platforms with a low computational capacity. However, a limitation is that the stored home location must be initially within the field of view of the robot. Motivated by the increasing ubiquity of camera information we propose to address this line-of-sight limitation by leveraging camera information from other robots and fixed cameras. To home to a location that is not initially within view, a robot must be able to identify a common visual landmark with another robot that can be used …

K-Means Clustering Using Gravity Distance, Ajinkya Vishwas Indulkar

Masters Theses & Specialist Projects

Clustering is an important topic in data modeling. K-means Clustering is a well-known partitional clustering algorithm, where a dataset is separated into groups sharing similar properties. Clustering an unbalanced dataset is a challenging problem in data modeling, where some group has a much larger number of data points than others. When a K-means clustering algorithm with Euclidean distance is applied to such data, the algorithm fails to form good clusters. The standard K-means tends to split data into smaller clusters during a clustering process evenly.

We propose a new K-means clustering algorithm to overcome the disadvantage by introducing a different …

The Causal Fairness Field Guide: Perspectives From Social And Formal Sciences, Alycia Carey, Xintao Wu

Computer Science and Computer Engineering Faculty Publications and Presentations

Over the past several years, multiple different methods to measure the causal fairness of machine learning models have been proposed. However, despite the growing number of publications and implementations, there is still a critical lack of literature that explains the interplay of causality-based fairness notions with the social sciences of philosophy, sociology, and law. We hope to remedy this issue by accumulating and expounding upon the thoughts and discussions of causality-based fairness notions produced by both social and formal (specifically machine learning) sciences in this field guide. In addition to giving the mathematical backgrounds of several popular causality-based fair machine …

A Low-Cost And Low-Tech Solution To Test For Variations Between Multiple Offline Programming Software Packages., Steffen Wendell Bolz

Masters Theses & Specialist Projects

This research paper chronicles the attempt to bring forth a low-cost and low-tech testing methodology whereby multiple offline programming (OLP) software packages’ generated programs may be compared when run on industrial robots. This research was initiated by the discovery that no real research exists to test between iterations of OLP software packages and that most research for positional accuracy and/or repeatability on industrial robots is expensive and technologically intensive. Despite this, many countries’ leaders are pushing for intensive digitalization of manufacturing and Small and Mediumsized Enterprises (SMEs) are noted to be lagging in adoption of such technologies. The research consisted …

A Low-Cost, Long-Range, And Solar-Based Iot Soil Quality Monitor, Salvador Garcia, Trina Nguyen, Julian Wong

Interdisciplinary Design Senior Theses

The project objective is to create a low-cost, long-range, and solar-based IoT soil quality monitoring system. The system must transmit packages of data gathered from separate nodes, consisting of two dierent types of sensors, to a centralized gateway receiver to be displayed to the user in an elegant and readable manner. The end goal of the project is to supplement produce grown by large agricultural bodies around the United States without the misuse of water resources. This report presents the need for this system, details the components of the system, and the rationale behind design choices. It serves as a …

Society Dilemma Of Computer Technology Management In Today's World, Iwasan D. Kejawa Ed.D

School of Computing: Faculty Publications

Abstract - Is it true that some of the inhabitants of the world’s today are still hesitant in using computers? Research has shown that today many people are still against the use of computers. Computer technology management can be said to be obliterated by security problems. Research shows that some people in society feel reluctant or afraid to use computers because of errors and exposure of their privacy and their sophistication, which sometimes are caused by computer hackers and malfunction of the computers. The dilemma of not utilizing computer technology at all or, to its utmost, by certain people in …

Cova Cci Undergrad Cyber Research, Nana Jeffrey

Cybersecurity Undergraduate Research Showcase

Is your digital assistant your worst enemy? Modern technology has impacted our lives in a positive way making tasks that were once time consuming become more convenient. For example a few years ago writing down your grocery list with a paper and pen was a norm, now with technology we have access to IoT devices such as smart fridges that can inform us on what items are low in stock, send a message to our digital assistants such as iOS Siri and Amazon's Alexa to remind us to buy those groceries. Although these digital assistants have helped make our daily …

Verifiable Searchable Encryption Framework Against Insider Keyword-Guessing Attack In Cloud Storage, Yinbin Miao, Robert H. Deng, Kim-Kwang Raymond Choo, Ximeng Liu, Hongwei Li

Research Collection School Of Computing and Information Systems

Searchable encryption (SE) allows cloud tenants to retrieve encrypted data while preserving data confidentiality securely. Many SE solutions have been designed to improve efficiency and security, but most of them are still susceptible to insider Keyword-Guessing Attacks (KGA), which implies that the internal attackers can guess the candidate keywords successfully in an off-line manner. Also in existing SE solutions, a semi-honest-but-curious cloud server may deliver incorrect search results by performing only a fraction of retrieval operations honestly (e.g., to save storage space). To address these two challenging issues, we first construct the basic Verifiable SE Framework (VSEF), which can withstand …

Towards Improved Inertial Navigation By Reducing Errors Using Deep Learning Methodology, Hua Chen, Tarek M. Taha, Vamsy P. Chodavarapu

Electrical and Computer Engineering Faculty Publications

Autonomous vehicles make use of an Inertial Navigation System (INS) as part of vehicular sensor fusion in many situations including GPS-denied environments such as dense urban places, multi-level parking structures, and areas with thick tree-coverage. The INS unit incorporates an Inertial Measurement Unit (IMU) to process the linear acceleration and angular velocity data to obtain orientation, position, and velocity information using mechanization equations. In this work, we describe a novel deep-learning-based methodology, using Convolutional Neural Networks (CNN), to reduce errors from MEMS IMU sensors. We develop a CNN-based approach that can learn from the responses of a particular inertial sensor …

Ux/U-Eye: Designing Graphical User Interfaces For Exclusive Eye Gaze Control, Timothy Curol

Honors Capstones

No abstract provided.

Generative Adversarial Networks Take On Hand Drawn Sketches: An Application To Louisiana Culture And Mardi Gras Fashion, Stephanie Hines

Honors Capstones

No abstract provided.

Development Of A Ppg Sensor Array As A Wearable Device For Monitoring Cardiovascular Metrics, Jose Ignacio Rodriguez-Labra

Masters Theses

Wearable devices with integrated sensors for tracking human vitals are widely used for a variety of applications, including exercise, wellness, and health monitoring. Photoplethysmography (PPG) sensors use pulse oximetry to measure pulse rate, cardiac cycle, oxygen saturation, and blood flow by passing a light beam of variable wavelength through the skin and measuring its reflection. A multi-channel PPG wearable system was developed to include multiple nodes of pulse oximeters, each capable of using different wavelengths of light. The system uses sensor fusion along with a machine learning model to perform feature extraction of relevant cardiovascular metrics across multiple pulse oximeters …

Discovering Ways To Increase Inclusivity For Dyslexic Students In Computing Education, Felicia Hellems, Sajal Bhatia

School of Computer Science & Engineering Faculty Publications

The years accompanying entrance into the university system are often characterized by a period of great transformation. These years can also be wrought with difficulties for many students, difficulties which are often compounded in students with disabilities (SWD). Reports from the U.S. Department of Education show that as recently as 2015--16, 19% of undergraduate students experienced some form of disability1. Additionally, statistics show that SWD tend to have lower post secondary completion rates than their counterparts [3]. A review of pertinent literature has shown that there still exist gaps within the field of computing education (CE) for teaching cybersecurity concepts …

An Overview Of The Potential For Blockchain Technology To Improve Cybersecurity, Stanley Mierzwa

Center for Cybersecurity

The purpose of this short research commentary is to provide a focused, semi-deep dive into the effort the industry places on cybersecurity defense and operations and the potential to integrate blockchain technology. As cybersecurity threats and incidents continue to rise, better procedures and strategies to protect our organizations’ data and systems are crucial to sustaining viable operations. Given that blockchain technology can potentially disrupt other industries (Moore, 2020), it is imperative to examine how it may improve our cybersecurity.

Cybersecurity Best Practices For The Manufacturing Industry, David Ortiz, Stanley Mierzwa

Center for Cybersecurity

The manufacturing and industrial sectors have evolved with the introduction of technologies over the past many decades. Progress in improving processes, techniques, output, quality, and efficiencies have been gained with new emerging technologies, resulting in positive and fortuitous changes for organizations. With the rapid movement towards a modern-day manufacturing environment, new and connected technologies that employ greater cyber-connectedness continue to grow, but at the same time, introduce cybersecurity risks.

Learning Domain Invariant Information To Enhance Presentation Attack Detection In Visible Face Recognition Systems, Jennifer Hamblin

Department of Electrical and Computer Engineering: Dissertations, Theses, and Student Research

Face signatures, including size, shape, texture, skin tone, eye color, appearance, and scars/marks, are widely used as discriminative, biometric information for access control. Despite recent advancements in facial recognition systems, presentation attacks on facial recognition systems have become increasingly sophisticated. The ability to detect presentation attacks or spoofing attempts is a pressing concern for the integrity, security, and trust of facial recognition systems. Multi-spectral imaging has been previously introduced as a way to improve presentation attack detection by utilizing sensors that are sensitive to different regions of the electromagnetic spectrum (e.g., visible, near infrared, long-wave infrared). Although multi-spectral presentation attack …

Assessing Security Risks With The Internet Of Things, Faith Mosemann

Senior Honors Theses

For my honors thesis I have decided to study the security risks associated with the Internet of Things (IoT) and possible ways to secure them. I will focus on how corporate, and individuals use IoT devices and the security risks that come with their implementation. In my research, I found out that IoT gadgets tend to go unnoticed as a checkpoint for vulnerability. For example, often personal IoT devices tend to have the default username and password issued from the factory that a hacker could easily find through Google. IoT devices need security just as much as computers or servers …

Pre-Training Graph Neural Networks For Link Prediction In Biomedical Networks, Yahui Long, Min Wu, Yong Liu, Yuan Fang, Chee Kong Kwoh, Jiawei Luo, Xiaoli Li

Research Collection School Of Computing and Information Systems

Motivation: Graphs or networks are widely utilized to model the interactions between different entities (e.g., proteins, drugs, etc) for biomedical applications. Predicting potential links in biomedical networks is important for understanding the pathological mechanisms of various complex human diseases, as well as screening compound targets for drug discovery. Graph neural networks (GNNs) have been designed for link prediction in various biomedical networks, which rely on the node features extracted from different data sources, e.g., sequence, structure and network data. However, it is challenging to effectively integrate these data sources and automatically extract features for different link prediction tasks. Results: In …

Dynamics And Simulations Of Discretized Caputo-Conformable Fractional-Order Lotka–Volterra Models, Yousef Feras, Semmar Billel, Al Nasr Kamal

Computer Science Faculty Research

In this article, a prey–predator system is considered in Caputo-conformable fractional-order derivatives. First, a discretization process, making use of the piecewise-constant approximation, is performed to secure discrete-time versions of the two fractional-order systems. Local dynamic behaviors of the two discretized fractional-order systems are investigated. Numerical simulations are executed to assert the outcome of the current work. Finally, a discussion is conducted to compare the impacts of the Caputo and conformable fractional derivatives on the discretized model.

The Music Bluetooth Controller: An Intersection Between Technology And Music, Lydia Wu

Senior Honors Theses

The modern musician faces a new challenge: how can technology be used to enhance a performance? This thesis documents the development of a Bluetooth remote controller that will aid today’s performing musicians by interacting with a digital display (e.g., an iPad) to flip musical score pages remotely. At its core, while mimicking a Bluetooth pedal (the current industry standard), this device attaches to the musician’s hand. In its pilot stages, the device has been referred to “MBC” (Music Bluetooth Controller).

Privacy-Preserving Information Security For The Energy Grid Of Things, Mohammed Alsaid, Nirupama Bulusu, Abdullah Barghouti, N. Sonali Fernando, John M. Acken, Tylor E. Slay, Robert B. Bass

Electrical and Computer Engineering Faculty Publications and Presentations

Smart grid infrastructure relies on information exchange between multiple actors in order to ensure system reliability. These actors include but are not limited to smart loads, grid control, and energy management technologies. As information exchange between these actors is susceptible to cyber-attacks, security and privacy issues are indispensable to ensure a reliable and stable grid. This position paper proposes a privacy-preserving, trust-augmented secure scheme for a smart grid implementation.

Quadratic Neural Network Architecture As Evaluated Relative To Conventional Neural Network Architecture, Reid Taylor

Senior Theses

Current work in the field of deep learning and neural networks revolves around several variations of the same mathematical model for associative learning. These variations, while significant and exceptionally applicable in the real world, fail to push the limits of modern computational prowess. This research does just that: by leveraging high order tensors in place of 2nd order tensors, quadratic neural networks can be developed and can allow for substantially more complex machine learning models which allow for self-interactions of collected and analyzed data. This research shows the theorization and development of mathematical model necessary for such an idea to …

Improving Feature Generalizability With Multitask Learning In Class Incremental Learning, Dong Ma, Chi Ian Tang, Cecilia Mascolo

Research Collection School Of Computing and Information Systems

Many deep learning applications, like keyword spotting [1], [2], require the incorporation of new concepts (classes) over time, referred to as Class Incremental Learning (CIL). The major challenge in CIL is catastrophic forgetting, i.e., preserving as much of the old knowledge as possible while learning new tasks. Various techniques, such as regularization, knowledge distillation, and the use of exemplars, have been proposed to resolve this issue. However, prior works primarily focus on the incremental learning step, while ignoring the optimization during the base model training. We hypothesise that a more transferable and generalizable feature representation from the base model would …