24067 research outputs found
Sort by
UC-1140 RiverGuard
No full textRiverGuard’s mission is to protect and preserve waterways by using technology to identify and reduce pollution. The system uses an object detection model to automatically locate and classify trash within images or video of rivers and lakes, removing the need for slow, manual observation. By providing real-time insight into waste accumulation, RiverGuard helps communities, researchers, and organizations take faster, more effective action to keep waterways clean. Its goal is to create a sustainable monitoring system that empowers people to understand pollution patterns and support long-term environmental responsibility. RiverGuard represents a step toward cleaner water, healthier ecosystems, and a more informed approach to environmental protection
GRP-20219 Continuous Monitoring of Cardiovascular Risk From Smartwatch Data Using a Knowledge Distillation Framework
No full textCardiovascular Disease (CVD) is one of the leading causes of global health concern, but current risk assessments are limited to episodic clinical visits. Most machine learning (ML) models trained on clinical data offer high accuracy but are not practical for continuous monitoring. Smartwatch-based wearables provide continuous real-time physiological data but lack clinical validation for robust risk prediction outside the clinical setting. To bridge this gap, we proposed a novel teacher-student knowledge distillation framework that transfers knowledge of complex and large EHR datasets to a small Fitbit smartwatch dataset-based prediction model. The student model achieves promising accuracy, identifying all types of derived CVD risk profile groups. Feature importance analysis revealed that daily average steps and sedentary time are the most prominent wearable-derived CVD risk predictors. Our study introduces a non-invasive continuous health monitoring framework, demonstrating that passively collected daily metrics can be transformed to clinically powerful early warnings of an individual’s long-term cardiovascular health plan
GRP-1231 Evaluating Generalization and Adaptation of Learning-Based Schedulers for Directed Acyclic Graph Workloads
No full textLearning-based schedulers such as Decima can optimize directed acyclic graph (DAG) workloads, yet their robustness under changing workload conditions is not well understood. This project evaluates how a Decima-trained policy transfers across different workload scenarios using an automated training and testing pipeline. Results show that the scheduler generalizes well to a workload with the same job scale, achieving a 1.9% improvement in average job completion time. Performance remains stable under a larger workload, but a shift in arrival pattern leads to an 83.7% increase in completion time and reduced fairness. These findings highlight both the potential and the limitations of learned scheduling policies, emphasizing the need for adaptive methods such as fine-tuning for reliable use in dynamic cluster environments
GRM-20242 CIPHER: Covert Influence Passed via Hidden Encoding in Representations Evaluating Subliminal Bias Transfer During Knowledge Distillation
No full textAI models can inherit hidden behavioral biases when student models learn from teacher outputs during knowledge distillation. Project CIPHER investigates whether covert signals, such as zero-width Unicode characters or column order shifts, can transmit bias from a teacher model to a student model even when the student never receives group labels. Using an experimental pipeline with controlled subliminal cues and dual distillation, the project aims to reproduce and measure subtle bias transfer. Preliminary results showed that weak signals produce no measurable bias, while the redesigned high-frequency signal and MLP student architecture reveal quantifiable disparity
UR-1163 Editing Films and Movies on GIMP
No full textThis project is to develop an import feature for Digital Picture Exchange (.DPX) files with GIMP. The DPX format is popularly used in the film and visual effects industry. DPX stores uncompressed high bit depth images. Although the project is still in progress, a working prototype is provided
GRM-20169 Proxy Recognition and Inclusive Scoring Method (PRISM): Evaluating Context-Dependent Bias in Large Language Models for Resume Screening
No full textAI-driven hiring tools are reshaping recruitment but often mirror biases in their training data. PRISM examines how large language models express or reduce demographic bias during resume evaluation and how linguistic context within prompts shapes these outcomes. Using a controlled dataset of 324 synthetic resumes with racially neutral surnames, differing only by first name as the demographic proxy, we compared GPT 3.5 turbo with a Sentence BERT similarity model. Under neutral prompts, no stable bias was observed across demographic groups, yet contextual shifts in the prompt changed how the model responded to proxy cues. These findings show that LLM bias can be either activated or suppressed depending on prompt framing, highlighting the significance of context-aware prompt design in improving fairness
Natural Disasters as Catalysts for Innovation in Architecture: Sustainable Solutions by Shigeru Ban and Nader Khalili
No full textJust as innovation is not limited to building types, compassion should not be limited by budget.
Natural and manmade disasters displace millions worldwide, creating urgent needs for affordable, safe, and dignified shelter. This paper explores how architects Shigeru Ban and Nader Khalili respond to these challenges by blending cultural context, using sustainable materials, and creative techniques Khalili’s earthbag structures, made from local earth and stabilized with minimal additives, offer economical, seismically safe homes inspired by traditional methods from Iran. Ban’s use of recyclable materials, such as paper and cardboard tubes, emphasizes emergency architecture in disaster zones like Japan and Korea, emphasizing fast, efficient use and environmental responsibility.
Through a comparative analysis of their work, this paper investigates how cultural context and material innovation influence humanitarian architectural design. The paper argues that both, winner Architects from the Prietzker Architecture Prize and Aga Khan Development Network Award, redefine disaster relief as an opportunity for creative, compassionate design that restores dignity and supports people recover.
This research demonstrates that effective disaster architecture does not have to rely on today\u27s high tech solutions but on culturally informed, resourceful answers that respond directly to human needs
Reimagining the Urban Neighborhood Library: A Sanctuary for Restoring Human Connection in the Tenderloin
No full textChildhood today, particularly ages 3-14, is increasingly defined by digital saturation and social disconnection, as screens and AI tools replace tactile, imaginative, and relational experiences that once nurtured growth and play. In this context, architecture can serve as a restorative force that rebuilds social connection and cognitive balance. This thesis explores how architecture can reimagine a neighborhood library in the Tenderloin as a family-centered civic sanctuary that accommodates children, fosters community interaction, and counteracts digital overstimulation, positioning it as a catalyst for the neighborhood’s development. The Tenderloin, San Francisco’s most child-dense yet open-space-deficient neighborhood provides an urgent testing ground where overstimulation, insecurity, and limited restorative environments erode daily well-being. Through a mixed-method research and design process, the study integrates psychology, neuroarchitecture, and biophilic design to identify architectural strategies that strengthen family interaction, encourage human connection, and transform civic space into an active framework for community life. Literature on restorative environments and precedents such as Helsinki’s Oodi and Charlotte’s ImaginOn inform design strategies of refuge, play, and adaptability. Site analysis of the Tenderloin examines demographic pressures, environmental stressors, and spatial inequities to anchor the project within real urban conditions. The resulting proposal envisions an urban neighborhood library that extends beyond its informational function to become restorative civic infrastructure: a family-engagement hub that weaves literacy zones, storytelling theaters, wellness rooms, and biophilic courtyards into an interconnected landscape of social and sensory relief. By positioning the library as both sanctuary and catalyst, the project demonstrates how architecture can transform digital overstimulation into an opportunity for urban renewal, family reconnection, and collective well-being.
Use of AI Disclaimer: Organization and language refinement
Dynamic Response and Vibration Analysis of Overhead Highway Traffic Sign Structures
No full textOverhead highway signage structures are essential components of transportation infrastructure, continuously exposed to dynamic aerodynamic loads generated by traffic and wind gusts. The cumulative effects of these cyclic loads, particularly from truck-induced wind pressures, can lead to fatigue cracking. Understanding their vibrational and dynamic behavior is therefore critical to ensuring long-term structural performance and safety.
This research investigates the vibration response of overhead sign structures through advanced computational methods. Building on previous analytical and experimental studies of Variable Message Sign (VMS) truss systems and steel hollow-section portal frames, this work reviews prior research conducted on two large-span VMS structures for The Alabama Department of Transportation (ALDOT) and introduces finite element modeling of similar structures in Georgia. Since no comparable research appears to have been conducted in Georgia, this study provides an approximate analysis and foundational reference for such structures in the state.
The study employs Finite Element Analysis (FEA) using Mecway software. Modal analysis was first performed to determine natural frequencies and vibration modes, followed by static analysis under wind-induced loading based on AASHTO guidelines. A parametric study was then carried out by varying member sizes and configurations to assess the influence of geometric parameters on stiffness and frequency response. The results contribute to understanding how design modifications can enhance system resilience and improve overall structural performance.
Beyond its technical focus, this project also serves an educational purpose by enriching undergraduate research experiences in structural engineering. It provides students with hands-on exposure to advanced FEA tools and dynamic analysis methods, strengthening their analytical and problem-solving skills while preparing them for graduate studies or professional practice in structural and research-oriented engineering.
Disclaimer: Portions of this abstract were refined using AI to improve grammar and conciseness. All scientific content, results, and conclusions were developed and verified by the authors
Designing a Green, Multistep Synthesis with an Intramolecular Diels-Alder Reaction for the Undergraduate Teaching Laboratory
No full textMultistep synthesis is a key learning objective for the undergraduate organic chemistry teaching laboratory. Yet, there is a notable gap in chemical education literature where multistep synthesis laboratory modules are embedded with green chemistry principles. The objective of this research endeavor is to develop a novel, multistep synthesis laboratory module that includes previously unexplored intramolecular Diels-Alder reactions. Our three-step synthesis utilizes bio-sourceable reagents—vanillin, furfurylamine, maleic anhydride—where each step is conducted under mild reaction conditions to enhance energy efficiency and minimize waste. The final product generated via an intramolecular Diels-Alder reaction reveals an unusual stereochemical outcome. This multistep synthesis confers high pedagogical value—allowing students to explore a stereospecific, intramolecular cycloaddition where each step is evaluated against the 12 Principles of Green Chemistry. In conclusion, this laboratory module offers a scalable model for sustainable synthetic practices in the undergraduate organic chemistry teaching laboratory