Other Computer Engineering Commons^™

Deep Learning-Based Change Detection In High-Resolution Remote Sensing Imagery, Hazem Badawy

Theses and Dissertations

Remote sensing has become a key tool for monitoring Earth’s surface over time, offering valuable insights into both natural and human-driven changes. Among its many applications, change detection focuses on analyzing multi-temporal imagery to reveal how specific areas evolve across different time periods. It plays a pivotal role in Earth observation applications, including urban development monitoring, environmental degradation assessment, and disaster response. However, existing approaches often struggle with limited contextual awareness, high sensitivity to noise, and imprecise localization of change boundaries, especially with high-resolution imagery. This thesis investigates the complex problem of change detection in remote sensing imagery by proposing …

Evaluation Of Deep Learning Techniques In Road Sign Recognition, Latafat Abbas Gardashova, Haji Fakhraddin Hajiyev

Chemical Technology, Control and Management

Deep learning has transformed the computer vision field and greatly improved the performance and efficiency of road sign recognition systems. This research compares different deep learning methods, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and hybrid models, in terms of their ability to effectively detect and classify road signs under various conditions. The study compares performance measures such as accuracy, processing speed, and robustness to environmental conditions like low lighting, occlusion, and adverse weather. The results show that CNN-based methods, especially those with transfer learning and ensemble techniques, have better performance in real-time scenarios. Problems like computational …

Adaptive Noise Estimation And Denoising With Deep Learning For Nmr Spectroscopy, Naveen Asokan

McKelvey School of Engineering Graduate Student Theses & Dissertations

Nuclear Magnetic Resonance (NMR) spectroscopy is a powerful analytical technique widely used for molecular structure elucidation in chemistry, biology, and medicine. However, spectral accuracy is often degraded by noise—particularly in low acquisition time settings—resulting in reduced resolution and obscured chemical features. While traditional noise reduction techniques such as signal averaging can improve spectral quality, they require longer acquisition times, limiting their utility in real-time and high-throughput applications.

This thesis presents a deep learning-based denoising framework designed to enhance the quality of complex-valued NMR spectra. The proposed model, built upon a U-Net architecture, incorporates both real and imaginary components of the …

Smart: Semantic Mapping And Analysis For Regional Terrain Using Multi-Scale U-Net And Topsis, Rashmi Sonth

Master's Projects

Accurate land use classification is the backbone for urban planning. But with poor quality satellite images, varied landscapes and structures which are changing faster than ever, it becomes a challenge to define clear boundaries and hence to urban planning. This research explores the application of deep-learning model for land use classification and asses the suitability of the land. The proposed model combines a multi-scale U-Net architecture with Transformer blocks applied on a multi-spectral satellite images that improves the semantic segmentation greatly across the urban and rural regions. Additionally, a patch-wise segmentation is applied to overcome the common problem of feature …

Learning From Leads: A 1d Dilated Resnet For Ecg Chagas Disease Screening, Somesh Saini, Matheus Lima Diniz Araujo

Student Scholarship

No abstract provided.

Augmenting Missing Sensor Data For Robust Human Activity Recognition, Suryakangeyan Kandasamy Gowdaman

Master's Projects

Applications of ubiquitous computing, including health monitoring, sports analytics, and ambient-assisted living, rely on Human Activity Recognition (HAR) using wearable sensors. However, model robustness is challenged by missing sensor values, class imbalance, inter-subject variability, and temporal noise. This work proposes a complete HAR pipeline that addresses these challenges through sampling, time-series augmentation, dynamic feature handling, and GAN-PCA-based imputation. Built on the DeepSense architecture, the model integrates convolutional feature extraction with bi-GRUs for temporal modeling. The system is evaluated using 5-fold cross-validation, subject-aware holdout, and LOSEO strategies on the Opportunity dataset. Results demonstrate consistent accuracy across folds and strong generalization to …

Suicidal Ideation Detection On Reddit Using Llm-Annotated Data And Graph Neural Networks, Ikbal Singh Gurdev Singh Dhanjal

Master's Projects

Suicide is the fourth leading cause of death among people aged 15-29. More than 720, 000 people commit suicide every year. During the COVID-19 pandemic, we saw an increase in people seeking out mental health support on anonymous forums like Reddit. These anonymous forums allow people to express their suicidal ideation without judgment and give them a support structure that not everyone has. The aim of this project is to detect suicidal ideation using Reddit. In this work, we propose SIRGEL (Suicidal Ideation on Reddit using Graph Embeddings and LLMs), a dual-pipeline approach that combines large language model (LLM)- based …

Coral Vision – Crustose Coralline Algae Detection With Computer Vision, Ryan Tseng

Master's Projects

Crustose coralline algae (CCA) are a group of red algae that are vital contributors to the health of coral reef ecosystems. Monitoring CCA abundance can serve as an indicator for coral reef health and improve reef conservation efforts. Autonomous Reef Monitoring Structures (ARMS) are artificial structures that can be deployed into coral reef ecosystems and retrieved to gather ecological data without harming reef structures. Traditional methods of calculating CCA abundance require manual analysis and are labor-intensive. Recent developments in computer vision and deep learning technology have provided an avenue to fully automate this task. This research aims to train a …

Implementation Of A Neural Network Execution Framework For Generalized And Cross-Platform Deep Learning Deployment And Inference On Spacecraft Systems, Rafael Polanco Segovia

Graduate Theses, Dissertations, and Problem Reports (ETD)

Recent advances in hardware and software technology have made it possible to implement more resource-demanding deep learning algorithms in constrained hardware environments. This creates opportunities to use deep learning for aerospace applications on increasingly smaller aerospace vehicles. This work presents the implementation of a Neural Network Execution Framework (NNEF), which aims to provide a cross-platform and reusable framework to deploy and execute trained neural networks for deep learning aerospace applications. The NNEF executes any neural network inference process regardless of the original deep learning framework in which it was created, for supported flight software platforms, and space-like computer boards. Users …

Designing Customized Loss Functions For Training Deep Neural Networks, Ali Pourramezan Fard

Electronic Theses and Dissertations

This dissertation explores the critical role of loss functions in enhancing the predictive performance of deep machine learning models. Loss functions are an integral element of all the ongoing advances we witness daily in this domain. I design custom loss functions and their impacts on various machine learning tasks, particularly in computer vision.

In the first stage of my research, I aim to improve the prediction performance of deep learning models by providing them with more precise feedback associated with task requirements. This led me to create the concept of assistive loss functions. My first proposed loss function, inspired by …

Impaired Speech Recognition Of Neurological Disorder Persons Using Machine Learning And Deep Learning Techniques, Vishnika Veni S

Theses and Dissertations

Speech Assistive Tools have emerged in recent years to support individuals with cognitive and neurological disorders in the field of assistive technology. People affected by neurological disorders such as autism, stroke, cerebral palsy, dysarthria, Parkinson’s disease, and brain injury often find it difficult to articulate desired sounds, resulting in impaired speech. As the population of impaired speakers continues to increase every year, there is a strong need to develop intelligent speech recognition systems for affected individuals. The primary objective of this research is to develop an Impaired Speech Recognition (ISR) system for the Tamil language. Word Recognition Accuracy (WRA) is …

M3t-Lm: A Multi-Modal Multi-Task Learning Model For Jointly Predicting Patient Length Of Stay And Mortality, Junde Chen, Qing Li, Feng Liu, Yuxin Wen

Engineering Faculty Articles and Research

Ensuring accurate predictions of inpatient length of stay (LoS) and mortality rates is essential for enhancing hospital service efficiency, particularly in light of the constraints posed by limited healthcare resources. Integrative analysis of heterogeneous clinic record data from different sources can hold great promise for improving the prognosis and diagnosis level of LoS and mortality. Currently, most existing studies solely focus on single data modality or tend to single-task learning, i.e., training LoS and mortality tasks separately. This limits the utilization of available multi-modal data and prevents the sharing of feature representations that could capture correlations between different tasks, ultimately …

Toward Intuitive 3d Interactions In Virtual Reality: A Deep Learning- Based Dual-Hand Gesture Recognition Approach, Trudi Di Qi, Franceli L. Cibrian, Meghna Raswan, Tyler Kay, Hector M. Camarillo-Abad, Yuxin Wen

Engineering Faculty Articles and Research

Dual-hand gesture recognition is crucial for intuitive 3D interactions in virtual reality (VR), allowing the user to interact with virtual objects naturally through gestures using both handheld controllers. While deep learning and sensor-based technology have proven effective in recognizing single-hand gestures for 3D interactions, research on dual-hand gesture recognition for VR interactions is still underexplored. In this work, we introduce CWT-CNN-TCN, a novel deep learning model that combines a 2D Convolution Neural Network (CNN) with Continuous Wavelet Transformation (CWT) and a Temporal Convolution Network (TCN). This model can simultaneously extract features from the time-frequency domain and capture long-term dependencies using …

Machine Learning Security For Tactical Operations, Dr. Denaria Fields, Shakiya A. Friend, Andrew Hermansen, Dr. Tugba Erpek, Dr. Yalin E. Sagduyu

Military Cyber Affairs

Deep learning finds rich applications in the tactical domain by learning from diverse data sources and performing difficult tasks to support mission-critical applications. However, deep learning models are susceptible to various attacks and exploits. In this paper, we first discuss application areas of deep learning in the tactical domain. Next, we present adversarial machine learning as an emerging attack vector and discuss the impact of adversarial attacks on the deep learning performance. Finally, we discuss potential defense methods that can be applied against these attacks.

Benchmarking And Enhancing Generalization In Multilingual Speech Emotion Recognition, Mohamed Osman Ismael

Theses and Dissertations

Speech Emotion Recognition (SER) is pivotal in advancing human-computer interaction by enabling machines to understand and respond to human emotions. Despite significant progress with self-supervised learning models, SER systems often struggle with generalization across diverse languages and unseen data distributions, limiting their real-world applicability. This thesis addresses these challenges by first introducing a large-scale benchmark to evaluate the robustness and adaptability of state-of-the-art SER models in both in-domain and out-of-domain settings. The benchmark includes a diverse set of multilingual datasets, emphasizing cross-lingual and out-of-domain evaluations to assess model generalization. Surprisingly, we find that the Whisper model, originally designed for automatic …

Comparitive Analysis Of Time Series Forecasting Using Frequency Informed Dense Neural Networks And Lstm, Avinash Mangalore Suresh

Master's Projects

Time series forecasting influences our lives on a daily basis, being a versatile tool in various application areas like environmental studies, finance, medicine and much more. While there are many established statistical and deep learning approaches to model time series data, each implementation comes with their own set of drawbacks or areas of improvements. Most of the existing deep learning architectures and research have focused on modeling time series data in the time-domain exclusively. However training deep learning models in the time-domain has some drawbacks, mainly due to the inherent temporal dependence of each time-step on the time-steps before it, …

Deep-Learning Approaches To Predict Remaining Useful Life Of Hard Disks, Rohan Mohapatra

Master's Projects

On a daily basis, data centers process huge volumes of data using inexpensive hard disks. Data stored in these disks serve a range of critical functional needs from financial, and healthcare to aerospace. As such, premature disk failure and consequent loss of data can be catastrophic. To mitigate the risk of failures, cloud storage providers perform condition-based monitoring and replace hard disks before they fail. By estimating the remaining useful life (RUL) of hard disk drives, one can predict the time-to-failure of a particular device and replace it at the right time, ensuring maximum utilization whilst reducing operational costs. We …

Novel Approach To Music Analysis Using Apache Spark, Nidhi Zare

Master's Projects

Music is one of the most common source of entertainment. Every user has their own taste of music and prefer to listen music that adheres to their taste and mood. There are various categories, called as music genres in which music can be classified. This research project addresses the challenge in music genre classification by using various deep learning models such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Very Deep Convolutional Networks (VGGNet), ResNet and others. The primary objective of this research is to enhance the accuracy of music genre classification using a distributed computing framework Apache Spark. …

Skin Cancer Detection Using Reinforcement Learning, Vikas Chercadu

Master's Projects

Advances in medical diagnostics have increasingly harnessed the power of artificial

intelligence, offering substantial improvements in early and accurate disease identifi- cation. This paper elaborates on a novel integration of Multi-Agent Reinforcement

Learning (MARL) with Deep Learning for the early detection of skin cancer, one of the most prevalent and lethal forms of cancer when left unchecked. Our project capitalizes on the sophisticated VGG16 Network for the extraction of detailed features from the widely-utilized HAM10000 dermatoscopic dataset, enhancing these features with additional color, texture, and shape analysis. Utilizing a custom-designed MARL environment, we facilitate a collaboration among multiple intelligent agents, …

Temporal Dynamics In Diabetes Prediction: A Sensor-Driven Time-Series Exploration, Monica Meduri

Master's Projects

Diabetes is a lifelong illness that, if not detected or managed appropriately, turns into serious complications. Correct glucose forecasting is critical to ensuring timely interventions, thereby minimizing risks of hyperglycemia and hypoglycemia, and optimizing the management strategies of the disease. Classical machine learning models have been applied in the blood glucose forecasting problem for a long time, however, usage of transformer-based architectures is still scarce within the literature. Due to the self-attention mechanism, transformers can capture temporal relationships very effectively, which makes them suitable for time-series data. TFT is a novel framework proposed here to utilize time-series data from CGM …

A Memory Efficient Deep Recurrent Q-Learning Approach For Autonomous Wildfire Surveillance, Jeremy A. Cantor

UNF Graduate Theses and Dissertations

Previous literature demonstrates that autonomous UAVs (unmanned aerial vehicles) have the po- tential to be utilized for wildfire surveillance. This advanced technology empowers firefighters by providing them with critical information, thereby facilitating more informed decision-making processes. This thesis applies deep Q-learning techniques to the problem of control policy design under the objective that the UAVs collectively identify the maximum number of locations that are under fire, assuming the UAVs can share their observations. The prohibitively large state space underlying the control policy motivates a neural network approximation, but prior work used only convolutional layers to extract spatial fire information from …

Detection Of Crypto-Ransomware Attack Using Deep Learning, Muna Jemal

Master of Science in Computer Science Theses

The number one threat to the digital world is the exponential increase in ransomware attacks. Ransomware is malware that prevents victims from accessing their resources by locking or encrypting the data until a ransom is paid. With individuals and businesses growing dependencies on technology and the Internet, researchers in the cyber security field are looking for different measures to prevent malicious attackers from having a successful campaign. A new ransomware variant is being introduced daily, thus behavior-based analysis of detecting ransomware attacks is more effective than the traditional static analysis. This paper proposes a multi-variant classification to detect ransomware I/O …

A Long-Term Funds Predictor Based On Deep Learning, Shuiyi Kuang

Electronic Theses, Projects, and Dissertations

Numerous neural network models have been created to predict the rise or fall of stocks since deep learning has gained popularity, and many of them have performed quite well. However, since the share market is hugely influenced by various policy changes or unexpected news, it is challenging for investors to use such short-term predictions as a guide. In this paper, we try to find a suitable long-term predictor for the funds market by testing different kinds of neural network models, including the Long Short-Term Memory(LSTM) model with different layers, the Gated Recurrent Units(GRU) model with different layers, and the combination …

Machine Learning Models To Automate Radiotherapy Structure Name Standardization, Priyankar Bose

Theses and Dissertations

Structure name standardization is a critical problem in Radiotherapy planning systems to correctly identify the various Organs-at-Risk, Planning Target Volumes and `Other' organs for monitoring present and future medications. Physicians often label anatomical structure sets in Digital Imaging and Communications in Medicine (DICOM) images with nonstandard random names. Hence, the standardization of these names for the Organs at Risk (OARs), Planning Target Volumes (PTVs), and `Other' organs is a vital problem. Prior works considered traditional machine learning approaches on structure sets with moderate success. We compare both traditional methods and deep neural network-based approaches on the multimodal vision-language prostate cancer …

Ensemble Transformer Architecture For Efficient And Real Time Sign Language Translation, Sumeet Ghegade

Master's Projects

Sign language is a form of visual language that uses face expression and hand gestures to communicate thoughts and concepts. The term refers to multiple visual languages that share some common visual cues but differ in their grammar and syntax. Sign language translation (SLT) is a crucial step in closing the communication gap between hearing and hearingimpaired people. The study of SLT using machine learning has gotten a lot of interest during the last three years but despite progress, SLT research is still in its early phases. Most of the previous approaches first convert the signs to glosses and then …

An Empirical Study Of Pre-Trained Model Reuse In The Hugging Face Deep Learning Model Registry, Wenxin Jiang, Nicholas Synovic, Matt Hyatt, Taylor R. Schorlemmer, Rohan Sethi, Yung-Hsiang Lu, George K. Thiruvathukal, James C. Davis

Department of Electrical and Computer Engineering Faculty Publications

Deep Neural Networks (DNNs) are being adopted as components in software systems. Creating and specializing DNNs from scratch has grown increasingly difficult as state-of-the-art architectures grow more complex. Following the path of traditional software engineering, machine learning engineers have begun to reuse large-scale pre-trained models (PTMs) and fine-tune these models for downstream tasks. Prior works have studied reuse practices for traditional software packages to guide software engineers towards better package maintenance and dependency management. We lack a similar foundation of knowledge to guide behaviors in pre-trained model ecosystems.

In this work, we present the first empirical investigation of PTM reuse. …

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 …

Deception Detection Across Domains, Languages And Modalities, Subhadarshi Panda

Dissertations, Theses, and Capstone Projects

With the increase of deception and misinformation especially in social media, it has become crucial to develop machine learning methods to automatically identify deception. In this dissertation, we identify key challenges underlying text-based deception detection in a cross-domain setting, where we do not have training data in the target domain. We analyze the differences between domains and as a result develop methods to improve cross-domain deception detection. We additionally develop approaches that take advantage of cross-lingual properties to support deception detection across languages. This involves the usage of either multilingual NLP models or translation models. Finally, to better understand multi-modal …

Generating Reality-Analogous Datasets For Autonomous Uav Navigation Using Digital Twin Areas, Thomas Lee, Susan Mckeever, Jane Courtney

Conference papers

In order for autonomously navigating Unmanned Air Vehicles(UAVs) to be implemented in day-to-day life, proof of safe operation will be necessary for all realistic navigation scenarios. For Deep Learning powered navigation protocols, this requirement is challenging to fulfil as the performance of a network is impacted by how much the test case deviates from data that the network was trained on. Though networks can generalise to manage multiple scenarios in the same task, they require additional data representing those cases which can be costly to gather. In this work, a solution to this data acquisition problem is suggested by way …

Digitalization Of Construction Project Requirements Using Natural Language Processing (Nlp) Techniques, Fahad Ul Hassan

All Dissertations

Contract documents are a critical legal component of a construction project that specify all wishes and expectations of the owner toward the design, construction, and handover of a project. A single contract package, especially of a design-build (DB) project, comprises hundreds of documents including thousands of requirements. Precise comprehension and management of the requirements are critical to ensure that all important explicit and implicit requirements of the project scope are captured, managed, and completed. Since requirements are mainly written in a natural human language, the current manual methods impose a significant burden on practitioners to process and restructure them into …