Computer Engineering Commons^™

Mosquito Classification And Explainability From Image Data Via Deep Learning Techniques, Farhat Binte Azam

USF Tampa Graduate Theses and Dissertations

According to the World Health Organization (WHO), mosquitoes are the deadliest animals on Earth, responsible for more human deaths annually than any other species. Mosquito-borne illnesses continue to pose severe risks to global health. In 2015 alone, there were an estimated 214 million malaria cases worldwide. Similarly, a 2016 report from the Centers for Disease Control and Prevention (CDC) revealed that Puerto Rico’s Department of Health received over 62,500 suspected cases of Zika, with 29,345 confirmed positive cases. In 2019, Southeast Asia experienced its worst dengue outbreak in recorded history. Of the approximately 4,500 mosquito species distributed across 34 genera, …

Edge-Streamer: A Lightweight, Self-Supervised Event Segmentation Model, Lucas Miller

USF Tampa Graduate Theses and Dissertations

Event segmentation is the practice of autonomously detecting the boundaries of semantically connected sequences of actions within a video. It is connected to many areas of computer vision, including action recognition and event understanding. Nearly all current work requires the entire video to be stored in memory to be processed in multiple passes. This takes up valuable resources, increases processing time, prevents the use of low-power, low-memory devices, and precludes long-form content and live videos from undergoing event segmentation.

This thesis introduces EDGE-STREAMER, a real-time, lightweight, self-supervised, transformer architecture capable of performing event segmentation in a single pass. EDGE-STREAMER uses …

Efficient Methods And Algorithms For Analyzing Stochastic Systems, Mohammad Ahmadi

USF Tampa Graduate Theses and Dissertations

This dissertation addresses the challenges of stochastic analysis of safety-critical systems with biological components, where unexpected behavior can lead to catastrophic events. Two fundamental challenges hinder the analysis of such systems: their typically large or infinite state spaces, and the extreme rarity of error states of interest. While Monte Carlo simulation can analyze biochemical systems without storing the state space, accurately estimating rare event probabilities becomes computationally prohibitive. Conversely, probabilistic model checking excels at analyzing extremely low probability events but becomes impractical for systems with large or infinite state spaces due to memory constraints.This work proposes two main contributions to …

Hierarchical Reinforcement Learning (Hrl) In Multi-Goal Spatial Navigation With Autonomous Mobile Robots, Brendon Johnson

USF Tampa Graduate Theses and Dissertations

Hierarchical reinforcement learning (HRL) is hypothesized to be able to take advantage of the inherent hierarchy in robot learning tasks with sparse reward schemes, in contrast to more traditional reinforcement learning algorithms. In this research, hierarchical reinforcement learning is evaluated and contrasted with standard reinforcement learning in complex navigation tasks. We evaluate unique characteristics of HRL, including their ability to create sub-goals and the termination function. We constructed experiments to test the differences between PPO and HRL, different ways of creating sub-goals, manual vs automatic sub-goal creation, and the effects of the frequency of termination on performance. These experiments highlight …

Bridging Virtual Robots And Physical Tasks Via Augmented Reality, Xiangfei Kong

USF Tampa Graduate Theses and Dissertations

Entry to human-robot interaction research, e.g., conducting empirical experiments, faces a significant economic barrier due to the high cost of physical robots, ranging from thousands to tens of thousands. This cost issue also severely limits the field’s ability to replicate user studies and reproduce the results to verify their reliability, thus offering more confidence to incorporate these findings. Although virtual reality (VR) user studies present a potential solution, it is unclear whether we can confidently transfer the findings to physical robots and physical environments because VR isolates both the physical robot and the physical world where robots operate. To address …

New Attack Surfaces Against Emerging Cloud And Web Based Infrastructures And Defenses, Junjie Xiong

USF Tampa Graduate Theses and Dissertations

Emerging network security threats, ranging from cloud-based infrastructure attacks to web-based content subversion, pose significant challenges to modern computing environments. In this dissertation, we explore two novel attack vectors that disrupt both cloud-based infrastructures and web-based content systems.

In this dissertation, we first introduce the Warmonger attack, a novel attack vector that can cause denial-of-service between a serverless computing platform and an external content server. The Warmonger attack exploits the fact that a serverless computing platform shares the same set of egress IPs among all serverless functions, which belong to different users, to access an external content server. As a …

Automated Data Analysis For Concussion Patient Records: A Flutter-Based Desktop Application, Fhaheem Tadamarry

USF Tampa Graduate Theses and Dissertations

Concussions are a prevalent and complex medical condition requiring careful clinical assessment and data-driven insights for effective management. This thesis presents the development of an automated data analysis system for concussion patient records, integrating Flutter-based desktop application development with SQL-driven data processing. The system provides a streamlined, interactive interface for clincians and researchers to upload, visualize, and analyze patient data efficiently.

The proposed solution automates data cleaning, preprocessing, and statistical analysis, ensuring robust and reliable insights into demographic, clinical, and recovery-related factors. Key analyses include sex-based differences injury mechanisms, prior head injury impact, mood disorder correlations, and time-to-treatment variations. The …

Robotic Multi-Object Grasping From A Pile: Techniques And Algorithms For Enhanced Dexterity, Tianze Chen

USF Tampa Graduate Theses and Dissertations

As robots become increasingly integrated into real-world applications such as warehousing, fulfillment centers, and manufacturing, the need for efficient and adaptable robotic systems grows. One of the key challenges is enabling robots to grasp multiple objects simultaneously, as this significantly boosts the efficiency of tasks like batch picking, sorting, and object transferring, reducing both time and energy consumption. This dissertation presents a comprehensive multi-object grasping (MOG) pipeline that includes pre-grasp selection, end-pose selection, grasping synergy calculation, and a data-driven model for estimating the number of objects being grasped. Central to this work is the development of the Experience Forest structure, …

Anonymized Identity Recognition And Classification Using Privacy Preserving Facial Encoding, Manas Sanjay Pakalapati

USF Tampa Graduate Theses and Dissertations

The need for sharing large-scale datasets to train deep neural network models, particularly in healthcare, raises significant data security and privacy concerns. To address these issues, methods such as data encryption or encoding are utilized. These techniques can encrypt the data and make it unreadable to humans, while still retaining its usefulness for training models.

In this study, we investigate various image encoding techniques designed to protect privacy by making images unrecognizable while still retaining their usefulness for model training. Our investigation utilized publicly available facial databases and focused on evaluating the trade-offs inherent in image encoding techniques, with a …

Specification, Enforcement, And Measurement Of Integrity Policies, Kevin Dennis

USF Tampa Graduate Theses and Dissertations

The first step to improving an organization's security posture is to define the organization's security goals. At a technical level, these goals are expressed as security policies. Security policies are predicates over programs, that return true or false if the program adheres to the policy. Defining these policies correctly is thus essential to ensuring the overarching security goals are met, but it is often quite difficult to translate human-oriented goals into their technical policy counterparts. In addition, these policies must be specified so that they are enforceable while minimizing false positives and false negatives. Integrity policies, which specify how data …

Human Motion-Inspired Inverse Kinematics Algorithm For A Robotics-Based Human Upper Body Model, Urvish Trivedi

USF Tampa Graduate Theses and Dissertations

The goal of this research is to develop a human motion-inspired inverse kinematics algorithm framework specifically designed for a Robotics-Based Human Upper Body Model (RHUBM). This framework offers solutions to challenges in various fields. In humanoid robotics, the framework addresses the problem of unnatural robot movement by enabling the development of motion planning algorithms that incorporate human-like movements. For prosthetics, the framework tackles the challenge of amputee difficulty in learning and controlling prosthetics by providing a user-friendly interface that predicts and visualizes upper limb movements, enabling learning and practice. In rehabilitation therapy, the framework tackles …

Exploring The Use Of Enhanced Swad Towards Building Learned Models That Generalize Better To Unseen Sources, Brandon M. Weinhofer

USF Tampa Graduate Theses and Dissertations

Deep learning models, typically, take significant time to train. Classifier ensembles are areliable way to increase classifier accuracy and perhaps generalizability to unseen sources of data. These classifiers can be combined with a simple voting scheme. The problem is that having multiple models can very heavily increase training time. Snapshot ensembles have been shown to provide a boost in performance by creating an ensemble of classifiers with different weights during the training of a single deep learned model. This can somewhat solve the problem of the increased training time as you do not have to train separate models. As Machine …

Programmable And Scalable Bit-Sliced Vlsi Architecture For Decision Tree Based Machine Learning Edge Inference, Raaga Sai Somesula

USF Tampa Graduate Theses and Dissertations

As the volume and diversity of Internet-of-Things (IoT) data continues to grow, traditional cloud-based processing methods face significant challenges, including latency, bandwidth constraints, and privacy concerns. Our research focuses on employing decision trees (DTs) as an intelligent filtering mechanism on the edge. Preliminary comparisons across four datasets revealed DTs are significantly more efficient than multilayer perceptrons (MLP), saving 97-98\% in area and power, leading to the selection of DT for our proposed architecture for lightweight IoT devices. We propose a novel programmable and scalable custom application specific integrated circuit (ASIC) architecture designed for Decision Tree based ML inference. Each bit-slice …

Under Pressure: The Soft Robotic Clap-And-Fling Of Cuvierina Atlantica, Daniel Mead

USF Tampa Graduate Theses and Dissertations

Evolution over billions of years has led to unique animals of all types. The tiny sea butterfly is one that remains mostly anonymous because of its size and low place on the food chain, but it is a swimming creature that when examined closely reveals a surprise. Shockingly, it has a swimming motion nearly identical to small insects flying at intermediate Reynolds numbers, referred to as the clap-and-fling. The centimeter long pteropod flapping briskly at 5Hz escapes attention with its size and speed. Modeling the clap-and-fling motion of the sea butterfly at a larger scale allows the benefits and uniqueness …

Brain-Inspired Spatio-Temporal Learning With Application To Robotics, Thiago André Ferreira Medeiros

USF Tampa Graduate Theses and Dissertations

The human brain still has many mysteries and one of them is how it encodes information. The following study intends to unravel at least one such mechanism. For this it will be demonstrated how a set of specialized neurons may use spatial and temporal information to encode information. These neurons, called Place Cells, become active when the animal enters a place in the environment, allowing it to build a cognitive map of the environment. In a recent paper by Scleidorovich et al. in 2022, it was demonstrated that it was possible to differentiate between two sequences of activations of a …

Machine Learning For Electronic Design Automation: Specification Mining And High-Level Synthesis, Md Rubel Ahmed

USF Tampa Graduate Theses and Dissertations

The rapid growth of complex system-on-chip (SoC) designs has presented unprecedented opportunities and challenges in electronic design automation (EDA). This dissertation explores two facets of electronic design automation: message flow specification mining using data mining and natural language processing (NLP) and high-level synthesis (HLS) acceleration using different machine learning (ML) methods. It also discusses an ML model co-optimization method for energy-efficient hardware implementation.

Effective SoC design validation relies heavily on message flow specifications. This dissertation presents an efficient technique for synthesizing finite state automaton (FSA) models from SoC execution traces. The synthesized models can provide valuable insights into the on-chip …

Deep Learning-Based Automatic Stereology For High- And Low-Magnification Images, Hunter Morera

USF Tampa Graduate Theses and Dissertations

Quantification of the true number of stained cells in specific brain regions is an important metric in many fields of biomedical research involving cell degeneration, cytotoxicology, cellular inflammation, and drug development for a wide range of neurological disorders and mental illnesses. Unbiased stereology is the current state-of-the-art method for collecting the cell count data from tissue sections. These studies require trained experts to manually focus through a z-stack of microscopy images and count (click) on a hundred or more cells per case, making this approach time consuming (~1 hour per case) and prone to human error (i.e., inter-rater variability). Thus, …

Iot Architecture For Enhancing Wayfinding And Accessibility In Smart Cities, Myles Keller

USF Tampa Graduate Theses and Dissertations

This research explores the potential of the Internet of Things (IoT) in enhancing urban wayfinding, particularly for the blind and visually impaired (BVI). Utilizing IoT's multi-layered architecture, smart cities deploy integrated sensors and devices to address urban challenges, with wayfinding as a key focus. While various accessible wayfinding methods have been explored, there exists a gap for more efficient solutions. Within this context, a collision warning system was developed for the BVI, integrated within a broader body sensor network (BSN) framework. Building on this, the study presents AWayNet, an RFID-based network, as a transformative solution. Simulations in the CARLA Simulator …

Analyzing Multi-Robot Leader-Follower Formations In Obstacle-Laden Environments, Zachary J. Hinnen

USF Tampa Graduate Theses and Dissertations

Observations in biological formation from nature likes flocks of birds, herds of mammals and packs of wolves have inspired the innovation of robotic architectures. This thesis presents an approach that aims to use robotic systems to mimic leader-follower behaviors in the navigation and formation of sparse and dense environments. The goal of this work is to extend and further analyze the original work of Weitzenfeld et al [3] to evaluate new swarm and pack based multi-robot architectures with the inclusion of obstacle avoidance and variations in group formations. The multiple robot architecture is based off a wolf pack with a …

Cyber-Physical Multi-Robot Systems In A Smart Factory: A Networked Ai Agents Approach, Zixiang Nie

USF Tampa Graduate Theses and Dissertations

This dissertation focuses on addressing the technical challenges of non-stationarity in smart factories through the use of cyber-physical AI agents. Industry 4.0 and smart manufacturing with smart factories as a central role, have a growing demand for Just-in-Time (JIT) and on-demand production, as well as mass customization—all while maintaining high productivity, resource efficiency and resilience. This research positions Multi-Robot Systems (MRS)-driven smart factories. The heterogeneous production and transportation robots in an MRS collaborate to form multiple real-time adjusted production flows achieving the flexibility to accommodate such on-demand, mass customization.

However, the implementation of MRS introduces new sets of challenges, including …

Enhancing And Securing Wireless Medical Technology For Diagnosis And Treatment Of Lower Urinary Tract Dysfunction, Farhath Zareen

USF Tampa Graduate Theses and Dissertations

Lower urinary tract dysfunction (LUTD) is a debilitating medical condition that affects millions of individuals worldwide. Urodynamics is the current gold standard for diagnosing LUTD but uses non-physiologically fast, retrograde cystometric filling to obtain a brief snapshot of bladder function. Current state-of-the-art research in bladder monitoring includes ambulatory urodynamics using wireless implantable devices to evaluate bladder function during natural filling for long-term monitoring. However, there are various challenges and limitations to this multi-sensor approach. This research focuses on developing frameworks for automated event detection, data analysis, and optimization of long-term bladder recordingsto improve the diagnosis and treatment of LUTD. In …

Optimization Techniques For Machine Learning Inference And Near Memory Image Processing In Hardware For Highly Constrained Iot Edge Nodes, Rajeev Joshi

USF Tampa Graduate Theses and Dissertations

The growing demand for fast and energy-efficient hardware for resource-constrained Internet of Things (IoT) edge devices has highlighted the limitations of conventional computing architectures. This research focuses on addressing the demand for fast, optimized, and energy-efficient machine learning inference engines as well as image processing in IoT edge applications. In this work, we address three challenging research problems and devise efficient solutions. Our investigation involved comprehensive exploration and analysis, leading to the proposal of effective approaches for overcoming these demanding issues. Through our work, we contribute novel solutions that offer improved efficiency and effectiveness in handling these research problems.

First, …

Secure Lightweight Cryptographic Hardware Constructions For Deeply Embedded Systems, Jasmin Kaur

USF Tampa Graduate Theses and Dissertations

Lightweight cryptography plays a vital role in securing resource-constrained deeply-embedded systems such as implantable and wearable medical devices, smart fabrics, smart homes, radio frequency identification tags, sensor networks, and privacy-constrained usage models. The National Institute of Standards and Technology (NIST) initiated a standardization process for lightweight cryptography, a relatively-long multi-year effort, which eventually concluded in February 2023. Side-channel attacks (SCAs) exploit the vulnerabilities of a system by observing and analyzing side-channel information leakages. Fault analysis attacks are a type of active SCAs, where an intelligent adversary injects bit/byte faults into the implementation of a cryptographic cipher to recover the secret …

Enhancing The Safety And Reliability Of Closed-Loop Medical Control Systems, Shakil Mahmud

USF Tampa Graduate Theses and Dissertations

The Internet of Medical Things (IoMT) is a rapidly advancing field that relies heavily on semi- or closed-loop Wearable and Implantable Medical Devices (WIMDs). In recent years, there has been renewed interest in clinical automation, with researchers looking for innovative solutions for Physiological Closed-Loop Control Systems (PCLCS). However, these devices can have various security issues, including vulnerabilities in software and firmware, physical attacks, weak encryption/authentication, and compromised system components. Government agencies, including the US Food and Drug Administration (FDA) and European Medicines Agency (EMA), emphasize the importance of ensuring the safety and reliability of PCLCS since malfunctioning medical devices can …

An Empirical Investigation Of Network Topology On Majority Illusion, Behavior Adoption, And Polarization, Sreeja Sreekantan Nair

USF Tampa Graduate Theses and Dissertations

Social influence plays a significant role in shaping opinions and behaviors both online and inthe physical world. The ways in which we interact with others and the information we receive can significantly influence our beliefs and attitudes towards different issues. While online social media provide platforms for such interactions, the underlying network structure can impact various social influence phenomena, including majority illusion, behavior adoption, and polarization.

This dissertation examines the role of network topology in three social influence processes:majority illusion, behavior adoption, and polarization. The majority illusion refers to a perception bias that occurs when individuals believe that an opinion …

Only Pick Once-Multi-Object Picking Algorithms For Picking Exact Number Of Objects Efficiently, Zihe Ye

USF Tampa Graduate Theses and Dissertations

Picking up multiple objects at once is a grasping skill that makes a human worker efficient in many domains. However, the State of Arts Robot Grasping skill has not developed such ability to compete with human.This paper presents a system to pick a requested number of objects by only picking once (OPO). The proposed Only-Pick-Once System (OPOS) contains several graph-based algorithms that convert the layout of objects into a graph, cluster nodes in the graph, rank and select candidate clusters based on their topology. OPOS also has a multi-object picking predictor based on a convolutional neural network for estimating how …

Insect Classification And Explainability From Image Data Via Deep Learning Techniques, Tanvir Hossain Bhuiyan

USF Tampa Graduate Theses and Dissertations

Since the dawn of the Industrial Revolution, humanity has always tried to make labor more efficient and automated, and this trend is only continuing in the modern digital age. With the advent of artificial intelligence (AI) techniques in the latter part of the 20th century, the speed and scale with which AI has been leveraged to automate tasks defy human imagination. Many people deeply entrenched in the technology field are genuinely intrigued and concerned about how AI may change many of the ways in which humans have been living for millennia. Only time will provide the answers. This dissertation is …

Secure Reconfigurable Computing Paradigms For The Next Generation Of Artificial Intelligence And Machine Learning Applications, Brooks Olney

USF Tampa Graduate Theses and Dissertations

The fields of artificial intelligence (AI) and machine learning (ML) have been popular tools for data analysis at the edge, particularly through complex deep and convolutional neural networks (DNNs/CNNs), which can learn to parameterize a function given a labeled dataset. Indeed, these technologies have enabled significant progress across a wide range of fields and are becoming ubiquitous. However, training the best model for an application and subsequently using them to evaluate data in real-time requires an immense amount of computational power. Typically, ``smart" sensors at the edge rely on the cloud to accelerate this computation due to power and compute …

A Human-In-The-Loop Robot Grasping System With Grasp Quality Refinement, Tian Tan

USF Tampa Graduate Theses and Dissertations

The goal of this dissertation is to develop a grasping system for assistive robots that can help people with disabilities and the elderly to perform tasks of daily living. In developing this robot grasping system, we maximize its reliability, accuracy, and autonomy. High reliability and accuracy are required for robots to perform tasks around human users and to safely interact with objects that might be fragile or have contents that could spill. High autonomy is desired as users with disabilities are usually not dexterous enough to directly operate the robot. In this dissertation, a human-in-the-loop (HitL) robot grasping system is …

Process Automation And Robotics Engineering For Industrial Processing Systems, Drake Stimpson

USF Tampa Graduate Theses and Dissertations

Automation in industrial systems applications has emerged as the fundamental solution for improving quality, production rate, and efficiency of a process. Much of the recent popularity surrounding the transition of processes from manually operated tasks to automated systems can be attributed to the concept of Industry 4.0, which outlines the fundamental guidelines for integrating cyber-physical systems into industrial processes. Due to rapid advancement of technology in robotics and automation as well as the increase in accessibility of resources to this technology, the capability to develop automated systems has become feasible for small-scale enterprise. This work presents a two-part initiative to …