Digital Commons Network^™

Instance-Level Video Depth In Groups Beyond Occlusions, Yuan Liang, Yang Zhou, Ziming Sun, Tianyi Xiang, Guiqing Li, Shengfeng He

Research Collection School Of Computing and Information Systems

Depth estimation in dynamic, multi-object scenes remains a major challenge, especially under severe occlusions. Existing monocular models, including foundation models, struggle with instance-wise depth consistency due to their reliance on global regression. We tackle this problem from two key aspects: data and methodology. First, we introduce the Group Instance Depth (GID) dataset, the first large-scale video depth dataset with instance-level annotations, featuring 101,500 frames from real-world activity scenes. GID bridges the gap between synthetic and real-world depth data by providing high-fidelity depth supervision for multi-object interactions. Second, we propose InstanceDepth, the first occlusion-aware depth estimation framework for multi-object environments. Our …

Efskip: A New Error Feedback With Linear Speedup For Compressed Federated Learning With Arbitrary Data Heterogeneity, Hongyan Bao, Pengwen Chen, Ying Sun, Zhize Li

Research Collection School Of Computing and Information Systems

Due to the communication bottleneck in distributed and decentralized federated learning applications, algorithms using compressed communication have attracted significant attention. The Error Feedback (EF) is a widely-studied compression framework for convergence with biased compressors such as top-k sparsification. Although various improvements have been obtained in recent years, the theoretical guarantee for EF-type framework is still limited. Previous works either 1) rely on strong assumptions such as bounded gradient/dissimilarity assumptions, thus can not deal with arbitrary data heterogeneity and also slow the convergence speed, or 2) can not enjoy linear speedup in the number of clients. In this work, we propose …

Scalable Graph Indexing Using Gpus For Approximate Nearest Neighbor Search, Zhonggen Li, Xiangyu Ke, Yifan Zhu, Bocheng Yu, Baihua Zheng, Yunjun Gao

Research Collection School Of Computing and Information Systems

Approximate nearest neighbor search (ANNS) in high-dimensional vector spaces has a wide range of real-world applications. Numerous methods have been proposed to handle ANNS efficiently, while graph-based indexes have gained prominence due to their high accuracy and efficiency. However, the indexing overhead of graph-based indexes remains substantial. With exponential growth in data volume and increasing demands for dynamic index adjustments, this overhead continues to escalate, posing a critical challenge.In this paper, we introduce Tagore, a fasT library accelerated by GPUs for graph indexing, which has powerful capabilities of constructing refinement-based graph indexes such as NSG and Vamana. We first introduce …

Llm-Assisted Cwe Identification, Severity Assessment, And Vulnerability Description Generation, Mohammad Jalili Torkamani

School of Computing: Dissertations, Theses, and Student Research

Identifying the underlying weakness types and assessing their severity using CWE and CVSS standards are critical steps in software vulnerability management. While timely assessment of vulnerabilities mitigates the impact of severe security incidents, automating joint CWE identification and severity assessment remains challenging due to the heterogeneity of vulnerabilities across different code granularities and programming languages. In addition, generating vulnerability descriptions is often time-consuming, as it requires extensive manual review, validation, and writing by security experts.

In this thesis, we leverage the capabilities of Large Language Models (LLMs) to automate the identification of CWE identifiers and the assessment of their severity …

Synthetic Dataset For Understanding Negation In Text-Guided Image Editing, Nhat-Tan Bui

Graduate Theses and Dissertations

Negation is a fundamental linguistic concept used by humans to convey information that they do not desire. Despite this, minimal research has focused on negation within text-guided image editing. This lack of research means that vision-language models (VLMs) for image editing may struggle to understand negation, implying that they struggle to provide accurate results. One barrier to achieving human-level intelligence is the lack of a standard collection by which research into negation can be evaluated. This thesis presents the first large-scale dataset, Negative Instruction (NeIn), for studying negation within instruction-based image editing. Our dataset comprises 366,957 quintuplets, i.e., source image, …

Programmable Network Approaches To Resilience And Security In Phasor Measurement Unit Networks, Zhiyao He

Graduate Theses and Dissertations

The security and resilience of smart grids are critical for ensuring reliable and stable power delivery. As modern power systems evolve to incorporate more advanced sensing and control capabilities, Phasor Measurement Units (PMUs) have become an important source of high-frequency, time-synchronized measurements that support wide-area monitoring, control, and protection. However, the growing complexity of smart grids and their reliance on real-time communication expose them to a range of cyber threats, including data loss, tampering, and coordinated attacks. This dissertation explores the use of programmable network technologies, particularly P4-based programmable switches, to provide in-network solutions that enhance the reliability and security …

Discriminative And Generative Video Modeling, Anh Pha Nguyen

Graduate Theses and Dissertations

Video modeling stands at the core of modern computer vision, enabling progress in domains such as surveillance, autonomous driving, and instructional assistance. Yet the complexity of spatiotemporal dynamics, multimodal integration, and the need for scalable and generalizable models present significant challenges. This dissertation addresses these issues from three complementary perspectives: discriminative modeling, multimodal (vision + language) alignment, and generative approaches, contributing new methods, datasets, and paradigms for advancing video understanding. In the discriminative setting, we propose a domain-adaptive framework for crowd counting that employs entropy minimization and adversarial learning to improve cross-domain generalization, and introduce a single-stage global association method …

Privacy Protection In Cloud-Based Biometric Systems, Yatish Reddy Dubasi

Graduate Theses and Dissertations

The widespread adoption of server-based biometric authentication systems, often hosted in the cloud, has introduced significant privacy risks. While these systems offer convenience, they require storing sensitive biometric templates on remote servers, creating a high-value target for adversaries. Unlike passwords, compromised biometric data is immutable and cannot be reissued, leading to an irreversible loss of privacy. This threat is exacerbated by template inversion attacks, which can reconstruct a user's original biometric trait (e.g., a face image) from its stored feature vector. This dissertation addresses these critical privacy challenges by designing, implementing, and evaluating a suite of novel frameworks for privacy-preserving …

Contextual Embedding Using Machine Learning For Cybersecurity: Access Control And Application, Thanh Bui

Graduate Theses and Dissertations

Access control is a well-established challenge in cybersecurity, with significant research focused on enhancing system autonomy and accuracy across various scenarios. Access control rules can be designed based on users’ roles, attributes, or relationships requesting access to specific resources. However, despite their benefits, these models still require human oversight. As systems expand and grow, it becomes increasingly complex for administrators to maintain precise access control rules, often necessitating extensive system updates or even a complete overhaul. This dissertation introduces a novel approach that leverages contextual embedding for user information to enable the system to autonomously authorize user requests for resources. …

When The Grid Goes Dark: A Digital Forensics Study Of Industrial Control System Cyberattacks, Katie Kettler

Graduate Theses and Dissertations

Industrial Control Systems (ICS) and Operational Technology (OT) maintain the grid, ensure water safety, and keep transportation running. Because they influence nearly every aspect of daily life, these systems have become prime targets for cyberattacks. The need for this research arises from the fact that when ICS and OT systems are compromised, the consequences go beyond data loss, and they can directly disrupt communities and endanger public safety. This thesis introduces digital forensics fundamentals and explains how investigations in ICS environments differ from those in traditional IT environments. This work then examines major attacks, including Stuxnet, the Ukrainian Grid Attacks …

Dancing With Logic: The Impact Of Integrating Dance In Teaching Introductory Computer Science Concepts On Student Understanding And Engagement, April Monk

Master's Theses

The purpose of this study was to investigate the effectiveness of dance-integrated pedagogical methods in enhancing the learning experiences of students in an introductory computer science course. In particular, the research aimed to measure the impact of creative movement on student comprehension, engagement, and perceptions of computer science. To guide this investigation, the study explored three research questions: Does integrating dance into lessons impact students’ comprehension of fundamental computer science concepts? How does dance integration affect student perceptions of both dance and computer science? And what elements of dance contribute to differences in comprehension between dance-integrated and traditional lessons? A …

Definition Of The 3d Position And Motion Status Of The Moving Heart Based On 2d Projections, Lawrence D. Orijuela

Electronic Theses, Projects, and Dissertations

This thesis presents a novel application of deep learning to the estimation of pulmonary vein coordinates using X-ray image pairs from a FORBILD Thorax phantom derived motion dataset. A Siamese neural network was developed to predict the 3D coordinates of one pulmonary vein at a time, specifically the Right Superior Pulmonary Vein (RSPV), Left Superior Pulmonary Vein (LSPV), Left Inferior Pulmonary Vein (LIPV), or Right Inferior Pulmonary Vein (RIPV), based on two-dimensional projection images.

The input data consisted of over 1.6 million grayscale X-ray image pairs across 1331 virtual patients, each annotated with ground truth 3D coordinates. To manage memory …

Few-Shot Anomaly Detection And Classification Through Reinforced Data Selection With A Combinatorial Reward, Xiaolei Hn, Depeng Xu, Shuhan Yuan, Xintao Wu

Electrical Engineering and Computer Science Faculty Publications and Presentations

Due to the scarcity of anomalies, deep anomaly detection models are typically trained in an unsupervised or semi-supervised manner, depending on the availability of a small number of labeled samples. Currently, most unsupervised approaches detect anomalies by identifying the deviant patterns from normal samples, and some semi-supervised studies also use labeled anomalies to improve performance. However, few studies have focused on how to take advantage of potential anomalies in an easily obtained and large-scale unlabeled dataset. Meanwhile, in a semi-supervised setting, although we assume there will be a small number of labeled anomalies, the task of anomaly classification is under-exploited, …

Marsanywhere: Dataset And Cross-View Diffusion Model For Satellite-To-Ground View Synthesis With Mars Data, Benjamin T. Hinchliff

Master's Theses

Satellite-to-ground view synthesis aims to create a realistic ground view image from a corresponding satellite view image. This is a well-studied problem for street level imagery, with good results being achieved by using modern image synthesis techniques such as diffusion models. However, despite the public availability of satellite and ground level imagery on Mars, these techniques have yet to be applied to the domain due to difficulties in collating and processing the data into a usable form. We address this deficiency by creating a dataset consisting of ground view panorama imagery from the Perseverance rover, along with associated satellite view …

Dual-Model Approach For Accurate Chest Disease Detection Using Gvit And Swin Transformer V2, Kamal Ahmad, Hafeez Ur Rehman, Babar Shah, Farman Ali, Irfan Hussain

All Works

The precise detection and localization of abnormalities in radiological images are very crucial for clinical diagnosis and treatment planning. To build reliable models, large and annotated datasets are required that contain disease labels and abnormality locations. Most of the time, radiologists face challenges in identifying and segmenting thoracic diseases such as COVID-19, Pneumonia, Tuberculosis, and lung cancer due to overlapping visual patterns in X-ray images. This study proposes a dual-model approach: Gated Vision Transformers (GViT) for classification and Swin Transformer V2 for segmentation and localization. GViT successfully identifies thoracic diseases that exhibit similar radiographic features, while Swin Transformer V2 maps …

Coupled Machine Learning Models: Combining Observations And Numerical Analysis In A Physics-Regularized Approach, Austin B. Schmidt

LSU New Orleans Theses and Dissertations

This dissertation investigates surrogate modeling for fixed-location environmental forecasting using novel data-combination techniques. The work surveys the landscape of observational measurements and numerically generated data, identifying similar research and gaps in current methodologies. The ratio-coupled training framework is introduced to combine two data sources per predicted feature through a tunable parameter that weights training signal strength. An optimization scheme is developed to simultaneously tune surrogate weights and the coupled signal ratio, allowing relative influence between signals to act as an explicit regularizer. Three case studies demonstrate the methodology and approach in a variety of contexts. The first study is based …

Topic Modeling And Culturomic Analysis Of 30,000 Books Over 100 Years Using Gensim, Michael A. Freeman

Electronic Theses and Dissertations

This thesis explores the cultural influence of historical events on English-language fiction published between 1820 and 1929. Using a corpus of 30,256 digitized books from Project Gutenberg, Latent Dirichlet Allocation (LDA) topic modeling was applied to identify recurring themes across eleven decades. The study sought to determine whether historically significant events could be detected within fictional narratives. One clear instance emerged: Napoleon Bonaparte and the Napoleonic Wars appeared explicitly in the 1820s corpus. Beyond this, several thematic patterns were observed—such as maritime language in the 1840s, national identity in the 1880s, and youth-oriented dialogue in the early 20th century—that plausibly …

Early Career Setback And Future Achievement In Professional Sports, Suman Kalyan Maity, Yang Wang, Nima Dehmamy, Victoria Medvec, Brian Uzzi, Dashun Wang

Computer Science Faculty Research & Creative Works

A central tenet of human performance posits that past success is a key predictor of future outcomes. This principle underpins selection processes in various human endeavors, shaping opportunity, wage, and winner-take-all inequalities. Here we systematically examine the future performance of previous winners and non-winners across two sports contexts using two different empirical strategies. First, we track young athletes participating in world-class track and field competitions and compare the future performance of bronze medalists to those finishing just shy of the podium. Next, we study a novel natural experiment in tennis, where we compare future performances of 'lucky losers'—players who advanced …

Understanding Bias And Fairness In Large Language Models: An Empirical Study, Joshua Johnson

Electrical Engineering and Computer Science Undergraduate Honors Theses

This thesis investigates demographic bias in large language models (LLMs) through the use of evaluating outcome disparities when utilized in decision making tasks as well as underlying associations that could contribute to furthering these disparities. Using profiles from the Adult dataset, we analyze how Gemini 2.0 Flash performs in an income prediction task using zero-shot and few-shot prompting methods. Our findings show that models exhibit measurable differences in demographic parity and false positive rates, with the use of few-shot prompting reducing these disparities. Alongside this line of testing, we tested associational bias in Qwen 2.5 using probability based association tests …

Security Vulnerabilities And Defense Tactics For Generative Ai Application Development, Kyle Klein

University Honors Theses

Generative AI (GenAI) applications such as OpenAI's ChatGPT leverage large language models (LLMs) trained on enormous amounts of data to accomplish tasks such as document editing, summarization, and query response. Chatbots and LLM programs that are equipped with retrieval-augmented generation (RAG) have the ability to draw upon data provided by developers and users to improve the quality of the program's responses. LLM technology has even expanded to generate images, audio, and video from user instructions. Designed around unpredictable user input and typically composed of many opaque components, LLM software products face a paradigm shift of new, constantly evolving security challenges. …

Content And Consequences: Impact Of Representation In Stem Higher Education Instructional Content On Marginalized Students, Nichole Ventura

Doctorate in Education

This qualitative study examined representation of historically marginalized students in STEM instructional content at the higher education level and its impact on their learning experiences. Despite growing diversity initiatives in STEM enrollment, curricular materials often fail to reflect the identities of underrepresented students. Using critical theory and interpretivist approaches, this research investigated how representation—or its absence—shapes students' sense of belonging, academic identity formation, and persistence. Through semi-structured interviews with undergraduate students from historically marginalized backgrounds, and purposeful sampling, this study captured the lived experiences of students engaging with STEM instructional materials. Interview protocols explored how students perceive their representation in …

A Software Framework For Translating Onnx Models Onto The Lace-C3a Hardware, Shawn Jones

All Graduate Theses and Dissertations, Fall 2023 to Present

Modern computers are powerful, but they are not always efficient enough for small, low power systems like those used on satellites and scientific instruments. To solve this problem, engineers often turn to FPGAs—reconfigurable computer chips that can be customized to run specific tasks much faster and with far less energy than ordinary processors. However, finding the best possible design for an FPGA program is extremely difficult because there are millions of ways a design could be built, and only a small fraction of them actually perform well.

This thesis presents SNOW, a new framework that helps automate the search for …

Upright-Net+: Enhanced Learning Of Upright Orientation For 3d Point Clouds, Xufang Pang, Feng Li, Hongjie Zhuang, Ning Ding, Xiaopin Zhong, Shengfeng He, Wenxi Liu, Bo Jiang

Research Collection School Of Computing and Information Systems

Automatic 3D shape analysis is heavily influenced by the pose of input 3D models, as the continuous nature of pose space introduces complexities that usually exceed the encoding capacities of standard deep learning frameworks. To tackle this challenge, we present Upright-Net+, an enhancement of our previous model, Upright-Net, specifically developed for estimating upright orientation in 3D point clouds. Our approach is grounded in the design principle that ”form ever follows function,” treating the natural base of an object as a functional structure that stabilizes it in its typical pose, influenced by physical laws and geometric properties. We reformulate the continuous …

Safe-Sora: Safe Text-To-Video Generation Via Graphical Watermarking, Zihan Su, Xuerui Qiu, Hongbin Xu, Tangyu Jiang, Jun-Hao Zhuang, Chun Yuan, Ming Li, Shengfeng He, Fei Yu

Research Collection School Of Computing and Information Systems

The explosive growth of generative video models has amplified the demand for reliable copyright preservation of AI-generated content. Despite its popularity in image synthesis, invisible generative watermarking remains largely underexplored in video generation. To address this gap, we propose Safe-Sora, the first framework to embed graphical watermarks directly into the video generation process. Motivated by the observation that watermarking performance is closely tied to the visual similarity between the watermark and cover content, we introduce a hierarchical coarse-to-fine adaptive matching mechanism. Specifically, the watermark image is divided into patches, each assigned to the most visually similar video frame, and further …

Iostom: Offline Imitation Learning From Observations Via State Transition Occupancy Matching, Quang Anh Pham, Brahmanage Janaka Chathuranga Thilakarathna, Tien Mai, Akshat Kumar

Research Collection School Of Computing and Information Systems

Offline Learning from Observations (LfO) focuses on enabling agents to imitate expert behavior using datasets that contain only expert state trajectories and separate transition data with suboptimal actions. This setting is both practical and critical in real-world scenarios where direct environment interaction or access to expert action labels is costly, risky, or infeasible. Most existing LfO methods attempt to solve this problem through state or state-action occupancy matching. They typically rely on pretraining a discriminator to differentiate between expert and non-expert states, which could introduce errors and instability—especially when the discriminator is poorly trained. While recent discriminator-free methods have emerged, …

Cssa-Fusion: Channel Selective And Spatial Alignment Infrared-Visible Image Fusion, Zhen Li, Zhi Zeng, Zhongrui Xiao, Ming Wen, Zhiyuan Zhang, Yibin Tian

Research Collection School Of Computing and Information Systems

Infrared-visible image fusion aims to integrate complementary information from two modalities to generate images with enriched semantic content. However, existing methods often neglect two critical aspects: the design of a local–global feature enhancement architecture and spatial alignment. To address these challenges, we propose Channel Selective and Spatial Alignment Fusion (CSSA-Fusion), a novel framework composed of two synergistic modules. The first is a selective channel and redundancy suppression module, which introduces a dual-branch selective channel attention mechanism to jointly capture local saliency and global channel importance for enhanced feature representation, and an informativeness–redundancy separation strategy to suppress redundant information while preserving …

The Rise Of Parameter Specialization For Knowledge Storage In Large Language Models, Yihuai Hong, Yiran Zhao, Wei Tang, Yang Deng, Yu Rong, Wenxuan Zhang

Research Collection School Of Computing and Information Systems

Over time, a growing wave of large language models from various series has been introduced to the community. Researchers are striving to maximize the performance of language models with constrained parameter sizes. However, from a microscopic perspective, there has been limited research on how to better store knowledge in model parameters, particularly within MLPs, to enable more effective utilization of this knowledge by the model. In this work, we analyze twenty publicly available open-source large language models to investigate the relationship between their strong performance and the way knowledge is stored in their corresponding MLP parameters. Our findings reveal that …

Contx: Scene Context Prediction Via Context Bank And Layout Perception, Jingxin Liang, Yangyang Xu, Haorui Song, Yuan Lu, Yuhui Deng, Yiyi Long, Yan Huang, Shengxin Liu, Jianbo Jiao, Shengfeng He

Research Collection School Of Computing and Information Systems

Scene context prediction, which seeks to infer unknown contextual information from isolated object properties, currently faces limitations due to predominant reliance on pixel-wise supervision that overlooks real-world context priors. To address this, we present ContX, a context-prior-driven, coarse-to-fine model. ContX distinctively integrates explicit linguistic-contextual knowledge in two key ways. First, it proposes a linguistic guided context bank, leveraging linguistic-statistical contextual data to guide the rationality of segmentation shapes and foster meaningful inter-class contextual interactions. Second, ContX augments contextual comprehension by correlating layouts with linguistic descriptions, enhancing layout perception through a multi-modal strategy. Comprehensive experiments demonstrate ContX's superiority and versatility, outperforming …

Pilot-C: Physics-Informed Low-Distortion Optimal Trajectory Compression, Kefei Wu, Baihua Zheng, Weiwei Sun

Research Collection School Of Computing and Information Systems

Location-aware devices continuously generate massive volumes of trajectory data, creating demand for efficient compression. Line simplification is a common solution but typically assumes 2D trajectories and ignores time synchronization and motion continuity. We propose PILOT-C, a novel trajectory compression framework that integrates frequency-domain physics modeling with error-bounded optimization. Unlike existing line simplification methods, PILOT-C supports trajectories in arbitrary dimensions, including 3D, by compressing each spatial axis independently. Evaluated on four real-world datasets, PILOT-C achieves superior performance across multiple dimensions. In terms of compression ratio, PILOT-C outperforms CISED-W, the current state-of-the-art SED-based line simplification algorithm, by an average of 19.2%. For …

General Test-Time Backdoor Detection In Split Neural Network-Based Vertical Federated Learning, Shunjie Yuan, Xinghua Li, Xuelin Cao, Haiyan Zhang, Robert H. Deng

Research Collection School Of Computing and Information Systems

As a new distributed machine learning framework, vertical federated learning (VFL) has been widely applied in the industry. However, recent studies have demonstrated that VFL faces serious challenges from backdoor attacks, which significantly hinder its further development. Although a few studies have focused on defending against VFL backdoor attacks, these defenses either do not consider the latest attack methods or show limited effectiveness. Moreover, most existing backdoor defense efforts primarily focus on backdoor attacks in horizontal federated learning (HFL) and centralized learning. Due to the unique architecture of VFL models, these methods cannot be directly applied to backdoor defense in …