Digital Commons Network^™

Portrait Shadow Removal Via Self-Exemplar Illumination Equalization, Qian Huang, Cheng Xu, Guiqing Li, Ziheng Wu, Shengxin Liu, Shengfeng He

Research Collection School Of Computing and Information Systems

We introduce the Self-Exemplar Illumination Equalization Network, designed specifically for effective portrait shadow removal. The core idea of our method is that partially shadowed portraits can find ideal exemplars within their non-shadowed facial regions. Rather than directly fusing two distinct classes of facial features, our approach utilizes non-shadowed regions as an illumination indicator to equalize the shadowed regions, generating deshadowed results without boundary-merging artifacts. Our network comprises cascaded Self-Exemplar Illumination Equalization Blocks (SExmBlock), each containing two modules: a self-exemplar feature matching module and a feature-level illumination rectification module. The former identifies and applies internal illumination exemplars to shadowed areas, producing …

Causality-Aware Safety Testing For Autonomous Driving Systems, Wenbing Tang, Mingfei Cheng, Renzhi Wang, Yuan Zhou, Chengwei Liu, Yang Liu, Zuohua Ding

Research Collection School Of Computing and Information Systems

Simulation-based testing is essential for evaluating the safety of Autonomous Driving Systems (ADSs). Comprehensive evaluation requires testing across diverse scenarios that can trigger various types of violations under different conditions. While existing methods typically focus on individual diversity metrics, such as input scenarios, ADS-generated motion commands, and system violations, they often fail to capture the complex interrelationships among these elements. For instance, identical motion commands can produce different collision risks in varying scenes, and the same collision may result from different commands under different scenarios. This oversight leads to gaps in testing coverage, potentially missing critical issues in the ADS …

Developing Blockchain-Based Transparent E-Commerce Solutions For Danish Smes To Promote Sustainable Design Products, Somnath Mazumdar, Robert John Kauffman, Thomas Jensen, Raghava Rao Mukkamala, Jan Damsgaard

Research Collection School Of Computing and Information Systems

Typically, a firm's objectives include establishing consumer confidence, preserving its brand image, and developing a profitable business strategy. Consumers now place greater emphasis on the sustainability and transparency of their purchases. Given environmental and economic limitations, firms are often compelled to implement sustainable production methods. This is especially a struggle for small- and medium-sized enterprises (SMEs) with new technology, as it can increase their risk of failure. This has led to a problem for consumers, who must cross-check the sustainability-related claims of the firms they buy from. This is challenging because of limited process trace data and restricted enforcement capabilities. …

Distributional Vision-Language Alignment By Cauchy-Schwarz Divergence, Wenzhe Yin, Zehao Xiao, Pan Zhou, Shujian Yu, Jiayi Shen, Jan-Jakob Sonke, Stratis Gavves

Research Collection School Of Computing and Information Systems

Vision-language alignment is crucial for various downstream tasks such as cross-modal generation and retrieval. Previous multimodal approaches like CLIP utilize InfoNCE to maximize mutual information, primarily aligning pairwise samples across modalities while overlooking distributional differences. In addition, InfoNCE has inherent conflict in terms of alignment and uniformity in multimodality, leading to suboptimal alignment with modality gaps. To overcome the limitations, we propose CS-Aligner, a novel framework that performs distributional vision-language alignment by integrating Cauchy-Schwarz (CS) divergence with mutual information. CS-Aligner captures both the global distribution information of each modality and the pairwise semantic relationships. We find that the CS divergence …

Managing Reproducibility Debt In Scientific Software: A Practical Framework, Zara Hassan, Christoph Treude, Graham Williams, Michael Norrish, Alex Potanin

Research Collection School Of Computing and Information Systems

Scientific software includes end-user applications, modelling tools, research software for publications, and production systems for real users. It plays a key role across various scientific disciplines by enabling large-scale computation, simulation, and data analysis. Unlike commercial software, scientific software is often developed in dynamic research environments with limited engineering practices, documentation, or testing. This makes it fragile and difficult to reproduce results, even when code and data are available, conditions in which Reproducibility Debt (RpD) accumulates. This paper presents the Reproducibility Debt Management Framework (RpD-MF), which is grounded in evidence from a systematic literature review, practitioner interviews, and a global …

Teamwise: Exploring Virtually Embodied Ai Facilitation For Video-Based Team Onboarding, Venkata Akhila Rani Obilisetty, Mikkeline Elleby, Anthony Tang, April Yi Wang

Research Collection School Of Computing and Information Systems

AI-mediated facilitation has emerged as a scalable approach to supporting onboarding and coordination in newly formed remote teams, yet existing systems are predominantly text-based. To explore how video-based, virtually embodied AI facilitators shape team experiences, we present TeamWise, which joins video-based onboarding meetings as an on-screen avatar. TeamWise guides teams through a structured facilitation flow of low-stakes activities to foster rapport, mutual awareness, and shared identity. While the overall sequence of activities and facilitation goals is predefined, the facilitator’s turn-by-turn utterances are generated dynamically by an LLM in response to participant input. We conducted a formative study of TeamWise to …

Patchgpt: Multi-Agent Patch Backporting Without Model Fine-Tuning, Ye Liu, Ruidong Han, Chengyan Ma, Yuqing Niu, David Lo

Research Collection School Of Computing and Information Systems

Patch backporting is crucial and prevalent in the maintenance of modern open-source software such as Linux kernels and forked repositories. However, porting patches across program versions remains a challenging problem due to the complexity of synergizing diverse patches with divergent program versions. In this paper, we propose PatchGPT, an agentic patch backporting framework for fine-grained patch generation. PatchGPT encompasses three agents: Miner for decomposing a sequence of atomic change steps as the original patch plan, Adapter for adapting the patch plan, and Executor for executing the adapted patch plan according to predefined change semantics. We conduct experiments on the PPatHF’s …

Understanding Codebase Like A Professional! Human-Ai Collaboration For Code Comprehension, Jie Gao, Yue Xue, Xiaofei Xie, Junming Cao, Soemin Thant, Erika Lee, Bowen Xu

Research Collection School Of Computing and Information Systems

Understanding an unfamiliar codebase is an essential task for developers in various scenarios, such as during the onboarding process. Especially when the codebase is large and time is limited, achieving a decent level of comprehension remains challenging for both experienced and novice developers, even with the assistance of large language models (LLMs). Existing studies have shown that LLMs often fail to support users in understanding code structures or to provide user-centered, adaptive, and dynamic assistance in real-world settings.To address this, we propose learning from the perspective of a unique role, code auditors, whose work often requires them to quickly familiarize …

Application Of The Tridiagonal Representation Approach And The J- Matrix Method Of Scattering In Theoretical Physics, Tunde Joseph Osunmusanmi

Dissertations

This dissertation is about the application of the Tridiagonal Representation Approach (TRA) in handling linear phenomenons, and for the first time, the J-matrix method of scattering in handling nonlinear phenomenons. The TRA is an algebraic method for solving linear ordinary differential equations of the second order. The advantage of the method in being algebraic is reinforced by the analytic power of orthogonal polynomials and special functions. On the computational side, it is favored as being reliant on powerful numerical techniques that deal with tridiagonal matrices such as Gauss quadrature and continued fraction. In the method, the solution of the differential …

Fully Decentralized Hierarchical Federated Learning At The Edge With Post-Quantum Secure Communication, Tariq Qayyum

Dissertations

Federated learning (FL) enables collaborative model training without centralizing raw data, but deploying FL at scale in real edge environments remains challenging because iterative training and aggregation must operate over heterogeneous, resource-constrained, and often mobile devices with time-varying connectivity. Conventional hierarchical federated learning (HFL) partially mitigates communication cost by introducing fog/edge aggregation, yet many designs retain cloud-based global aggregation and cloud-centric coordination. This places wide-area network latency on the critical path of every training round, creates a single point of failure, and limits responsiveness as model sizes and federation scale grow. Moreover, moving coordination and aggregation closer to the edge …

Ideals And Lattices In Number Fields, Sarah Ali Alyammahi

Theses

This thesis investigates algebraic number fields and their rings of integers, which

generalize the ring of integers ℤ in ℚ. The study focuses on ideals, units, and ideal class

groups, which describe the arithmetic structure of number fields and the failure of unique factorization. Key invariants such as the norm, trace, and discriminant are developed and applied, with particular emphasis on quadratic number fields and classical examples such as the Gaussian and Eisenstein integers. Some explicit computations of ideal class groups are carried out. The thesis also explores connections with lattice theory by interpreting rings of integers as lattices and …

Traditional And Machine-Learning Equalization Techniques For Bandwidth-Limited Short-Reach Optical Communication Channels, Abdullah Khawatmi

Theses

This thesis investigates equalization techniques for bandwidth-limited short-reach optical communication systems, with a focus on Visible Light Communication (VLC) and Step-Index Plastic Optical Fiber (SI-POF) links. Commercial light-emitting diodes and photodiode receivers impose severe bandwidth constraints, inter-symbol interference, and noise sensitivity, which fundamentally limit achievable data rates. The work addresses these impairments through systematic evaluation of traditional digital signal processing–based equalizers and modern machine-learning-based post-equalization methods. The primary aim of this thesis is to enhance the achievable data rate and reliability of commercial short-reach optical links while maintaining practical computational complexity. Specifically, the objectives are to (i) design and experimentally …

Intelligent Deep Learning-Based Sign Language Translation System, Nada Rasem Shahin

Dissertations

The Deaf and Hard of Hearing (DHH) community uses sign language as a primary means of communication. However, the shortage of sign language interpreters and the existence of hundreds of sign languages limit accessibility and inclusion. Sign Language Machine Translation (SLMT) systems present a promising solution for bridging the communication gap between the DHH and the hearing individuals, supporting inclusive societies. In smart cities, such systems play an essential role in improving the quality of life on a community level. In particular, as the population’s well-being is critical, developing intelligent assistive technologies, such as SLMT systems, is necessary to provide …

Efficient Energy Management In Networked Microgrids Using Multi-Agent Deep Reinforcement Learning In The Presence Of Uncertainties, Ayodele Benjamin Chukwuyem

Dissertations

Microgrid technology is essential in facilitating the transition to smart energy grids in developed countries and mitigating energy poverty in developing countries, particularly in areas where grid extensions are not feasible. Recently, the concept of networked microgrids (NMGs) has garnered tremendous attention due to the plausibility of interactions among interconnected microgrids leading to power networks that are more resilient, reliable, and stable. However, because each microgrid has diverse distributed generation resources (renewables and controllable generators) and each microgrid operator (MO) has different objectives, coordinated energy management is required to satisfy local and system-wide goals under conditions with significant uncertainty. Existing …

Utilizing And Optimizing Forecasting Models For Nursing Demand: A Narrative Review, Kalpana Singh, Liyan Ajit D. Souza, Ananth Nazarene, Amane Mounia, Mahmoud Abdelwahab Khedr, Ahmad Mousa, Rebecca Rejo George, Abdulqadir J. Nahwan

Research outputs 2022 to 2026

Background: Accurately forecasting nursing demand is essential for effective workforce planning in the context of increasing patient volumes and the growing complexity of healthcare systems. Reliable forecasting supports optimal staffing, enhances patient care quality, and reduces operational risks. However, traditional forecasting approaches are increasingly challenged by evolving clinical demands, rising chronic disease burden, and rapid technological advancements. Aim: This narrative review aimed to (1) map existing models used for predicting nursing demand, (2) identify key predictive methodologies, data inputs, and evaluation metrics, and (3) highlight existing gaps and implications for healthcare policy and workforce management. Methods: A comprehensive narrative literature …

Dendrimer Nanogels With Built-In Free Radical Scavenging Enable Efficient Topical Delivery Of A Hydrophilic Antioxidant To Restore Lens Redox Balance For Cataract Treatment, Lin Qi, Huari Kou, Anna Chernatynskaya, Da Huang, Vimalin Jeyalatha Mani, Humeyra Karacal, Nuran Ercal, Hu Yang

Chemistry Faculty Research & Creative Works

Cataract is a leading cause of vision impairment worldwide and are primarily caused by oxidative stress that damages and aggregates lens proteins, leading to lens opacification. However, the eye's anatomical barriers limit the penetration and bioavailability of antioxidant therapies. To address this challenge, a dendrimer-based nanogel with a built-in reactive oxygen species (ROS)-scavenging capability developed by us was employed to deliver the antioxidant N-acetylcysteine (NAC) to the lens. NAC was loaded into a generation-5 PEGylated poly(amidoamine) dendrimer (G5-PEG-TK, termed the GPT) nanogel. The resulting NAC-GPT was characterized for its ROS-scavenging activity, bioavailability, and corneal permeability. The efficacy of NAC-GPT was …

Addressing Class And Demographic Imbalance In E-Commerce Behavior Prediction: A Case Study Using Resampling Techniques, Nurul Ain Mustakim, Maslina Abdul Aziz, Shuzlina Abdul Rahman, Rahmiati Rahmiati

Malaysian Journal of Computing (MJoC)

In e-commerce predictive modeling, imbalanced data remains a critical challenge, particularly when both class labels and demographic attributes are unequally distributed. This study investigates a combined approach of Synthetic Minority Oversampling Technique (SMOTE) and demographic resampling to improve the performance of models predicting online purchasing behavior in Malaysia. Using a dataset of 1,126 survey responses, six classifiers (J48, Random Tree, REPTree, JRip, PART, and OneR) were evaluated under three conditions: unbalanced, after SMOTE, and after SMOTE with demographic balancing. The results displayed clear improvements in model performance. For example, J48’s accuracy increased from 62.85% (unbalanced) to 98.69% (fully balanced), while …

Textual Adversarial Example Generation Using Bigram Unigram-Semantic Preservation Optimization Algorithm, Noor Adam Noor Azmi, Haslizatul Fairuz Mohamed Hanum

Malaysian Journal of Computing (MJoC)

The vulnerability of Natural Language Processing (NLP) models to adversarial attacks remains a critical challenge in the field of cybersecurity and AI robustness. While deep learning models have achieved high performance in sentiment analysis, they are susceptible to subtle input perturbations that induce misclassification. This study presents the design and practical implementation of a web-based system (Proof of Concept) that automates the generation of textual adversarial examples using the Bigram Unigram-Semantic Preservation Optimization (BU-SPOF) algorithm. Rather than proposing a novel attack algorithm, our primary contribution is the architectural integration of a dual-source candidate generation strategy (WordNet and OpenHowNet) and a …

Expanding Traditional Literacy Frameworks: A Multiliteracies Approach In Elementary Classrooms, Caroline Cuenca

Doctoral Dissertations

Teaching literacy in elementary schools can provide students with a rich, meaningful, and engaging learning experience. Furthermore, cultivating a culturally responsive approach to teaching and learning can promote feelings of acceptance and belonging in young children (Comber, 2018). However, elementary educators are feeling increased pressure to emphasize literacy instruction with a sole focus on foundational literacy skills, standardized curriculum, and accountability measures (Unadkat & Ochoa, 2026). Unfortunately, these instructional practices often limit students' opportunities to engage in creative, collaborative, multimodal, and meaningful learning (Unadkat & Cuenca, 2025). This mixed-methods study examined how a critical multiliteracies curriculum was implemented in a …

Multiscale Effects Of Landscape Structure, Biodiversity, And Chronic Physiological Stress In Atlantic Forest Small Mammals, Lizette Arroyo

Theses and Dissertations from DePaul University

Habitat loss and fragmentation are major drivers of biodiversity decline and can alter the physiological condition of wildlife through changes in habitat quality, resource availability, and community composition. The Atlantic Forest of South America is one of the world’s most threatened biodiversity hotspots. This study examined relationships among biodiversity, landscape structure, and chronic stress in nonvolant small mammals within the Reserva Natural Tapytá in eastern Paraguay using hair glucocorticoid concentrations as indicators of long-term physiological stress. Small mammals were sampled from forest fragments representing multiple patch sizes between 2023 and 2025. Hair cortisol and corticosterone concentrations were quantified using enzyme …

Grading Machines: Can Ai Exam-Grading Replace Law Professors?, Kevin L. Cope, Jen Frankenreiter, Scott Hirst, Eric A. Posner, Daniel Schwarcz, Dane Thorley

Faculty Scholarship

In the past few years, large language models (LLMs) have achieved significant technical advances, such that legal-advocacy organizations are increasingly adopting them as complements to—or substitutes for—lawyers and other human experts. Several studies have examined LLMs' performance in taking law school exams, finding mixed results. Yet there have been no published studies systematically analyzing LLMs' competence at one of law professors' chief responsibilities: grading law school exams. This paper presents results of an analysis of how LLMs perform in evaluating student responses to legal analysis questions of the kind typically administered in law school exams. The underlying data come from …

Towards Reliable And Trustworthy Deep Learning Through Explainability And Interpretability, Dipkamal Bhusal

Theses

Deep neural networks achieve state-of-the-art performance across many domains, yet their deployment in high-stakes settings is constrained by two challenges: opaque decision-making and vulnerability to adversarial manipulation. This thesis investigates explainability and interpretability as principled mechanisms for improving the reliability and trustworthiness of deep learning models. First, we develop new post-hoc explanation methods that improve feature attribution and concept-based explanations. These methods provide faithful decision cues by modeling meaningful feature interactions and extracting faithful coherent concepts, enabling more reliable understanding of why a model predicts a given label. Second, we show that explanation quality is not solely a property of …

A Comparative Study Of Inference-Time Scaling Strategies For Large Language Models, Oluwamayowa Owolabi

Theses

Large language models (LLMs) have demonstrated strong performance on a range of reasoning tasks, however, their reliability often depends not only on model size or training data, but also on inference-time strategies. However, existing inference-time methods are typically evaluated in isolation and under differing experimental assumptions, making it difficult to draw systematic conclusions about their relative effectiveness. This thesis proposes a controlled empirical study of inference-time scaling strategies for large language models under fixed inference-time compute budgets. The findings reveal that no single strategy dominates uniformly. PRM guided selection with the IBM Granite verifier achieves the highest absolute accuracy across …

Software Vulnerability Recidivism In Open-Source Projects, Brandon Keller

Theses

Software vulnerabilities present a major threat to businesses and individuals alike and it is therefore critical that a culture exists among software engineers to encourage the discovery and patching of security flaws. Vulnerability counts are a common way of evaluating a project’s security. However, this metric can run counter to building a developer culture of fault recognition if more vulnerabilities is always seen as a bad thing. While these counts can present a rough idea of a project’s history with security, they provide no insight into how the development team improves and learns as a result of a vulnerability. A …

Effect Of Geometry, Cell Size, And Carbon Fiber Reinforcement On The Charpy Impact Strength Of Additively Manufactured Tpms Lattice Structures, Ahmed Yousuf Mohammad Bin Yaroof (Alsuwaidi)

Theses

Scientific research indicates a growing utilization of lightweight, high-performance materials across various disciplines, including engineering, driven by the demands of modern technological advancements. Recent developments within these fields include the creation of advanced lattice structures through additive manufacturing (AM) processes. One such lattice structure is the triply periodic minimal surface (TPMS) structure. TPMS structures possess unique mechanical properties and energy absorption characteristics that distinguish them from conventional AM lattice structures. However, published research to date has largely focused on the quasi-static behavior of TPMS structures, with limited attention given to the impact performance of composite- reinforced TPMS structures. The purpose …

Preparing Future-Ready Graduates For Technological And Workforce Transformations, Ghia El Dirani

Theses

The United Arab Emirates (UAE) is undergoing rapid economic transformation driven by technolog- ical innovation and national strategies such as Vision 2031 and UAE Centennial 2071, positioning STEM (Science, Technology, Engineering, and Mathematics) education as critical to building a knowledge-based economy. However, a persistent gap exists between the competencies developed in STEM higher education programs and the skills demanded by emerging sectors such as artificial intelligence, renewable energy, and advanced manufacturing. While the UAE has introduced pro- gressive education policies and invested in digital infrastructure, most curriculum reforms remain reactive and disconnected from long-term workforce projections. This research applies strategic …

Investigation Of Machining (Drilling) Of Bio-Composite Reinforced With Jute Fibers Under Different Machining Conditions, Mohammed Abdul Mujeeb Ansari

Theses

This study investigates the drilling performance of jute fibre reinforced bio-composites under different machining conditions to reduce performance parameters like surface roughness, tool type and delamination damage (entry and exit). Jute composites were fabricated using a process called hand layup. Drilling experiments were carried out using full factorial method focusing on three parameters: Feed Rate (0.05mm/rev, 0.1mm/rev and 0.15mm/rev), Drill bit Type (Titanium Nitride-TiN, High Speed Steel-HSS and High-Speed Steel Cobalt HSS-Co) and Lubrication condition (Dry, Minimum Quantity Lubrication-MQL and Cryogenic-LN2). A total of 27 experimental trials were performed, and the responses were measured for each condition. Statistical analysis was …

Advanced Artificial Intelligence Vs Simpler Models For 1-Year Death Prediction Among Patients Receiving Hemodialysis, Karthikeyan K, Jennifer E Flythe, Patrick H Pun, Wolfgang C Winkelmayer, David Carlson

Faculty, Staff and Students Publications

Objectives: We evaluated the data requirement for modern AI tools to outperform simpler models in predicting short-term mortality in over 500 000 patients with hemodialysis-dependent kidney failure.

Materials and methods: We compared logistic regression, boosting, and transformers using increasingly complex feature sets (from last-visit data to full trajectories). Performance was measured using the area under the ROC curve (AUC-ROC) and the Precision-Recall curve (AUC-PR) across training data sizes ranging from 500 to 490 197 samples.

Results: Using features with temporal information is beneficial across all models. On the full dataset, Transformers (AUC-ROC = 0.8568) and boosting (AUC-ROC = 0.8598) perform …

Fronto-Cerebellar Features Associate With Cognitive Dysfunction In Childhood-Onset Systemic Lupus Erythematosus, Hanne Van Der Heijden, Gabrielle Alonzi, Amanda Cao, Raquel Van Gool, Merve Koç Yekedüz, Lise Vrolix, Itamar Ronen, Vanessa Rameh, Kyle Mcbrearty, Aditi Deokar, Robert P Sundel, Eyal Muscal, Joseph Gonzalez-Heydrich, Andrea Knight, Joyce C Chang, Jaymin Upadhyay

Faculty, Staff and Students Publications

Objective: Cognitive dysfunction (CD) is a prevalent symptom in childhood-onset systemic lupus erythematosus (cSLE). This study aimed to investigate the neurobehavioral basis of CD in cSLE.

Methods: Patients with cSLE (N=20) and age- and sex-matched healthy controls (HCs, N=20) completed questionnaires and multiple neurocognitive tests. The Systemic Lupus Erythematosus Disease Activity Index 2000 and laboratory markers were used to monitor patients' clinical status. Neuroimaging assessments included functional near-infrared spectroscopy (fNIRS), functional magnetic resonance imaging (fMRI), and structural MRI.

Results: cSLE patients demonstrated moderate disease activity with high inflammation and immune dysregulation, alongside low medication adherence. Relative to HCs, cSLE patients …

Toward A Unified Framework For Open World Visual Learning, Yuansheng Zhu

Theses

Artificial intelligence systems have achieved remarkable performance across a wide range of visual tasks. However, most existing models operate under the unrealistic closed-world assumption, where training and test data are drawn from the same distribution. In real-world applications such as anomaly detection, autonomous driving, and medical diagnosis, learning systems frequently encounter novel or out-of-distribution scenarios. These settings require models that can recognize unknown inputs, adapt to new information over time, and maintain reliable performance under evolving conditions. This dissertation studies the problem of Open World Visual Learning, a paradigm that enables visual learning systems to operate robustly in dynamic and …