UTSA Runner Research Press (Univ. of Texas at San Antonio)
Not a member yet
6846 research outputs found
Sort by
Improving Resource and Energy Efficiency for Cloud 3D through Excessive Rendering Reduction
No full textThe rise of cloud gaming makes interactive 3D applications an emerging type of data center workload. However, the excessive rendering in current cloud 3D systems leads to large gaps between the cloud and client frame rates (FPS, frames per second), thus wasting resources and power. Although FPS regulation can remove excessive rendering, due to the highly-varying frame processing time and the use of rendering delays, existing cloud FPS regulation solutions have low FPS and slow motion-to-photon (MtP) latency, causing violations of Quality-of-Service (QoS) requirements.
In this paper, we present a novel cloud FPS regulation solution, called OnDemand Rendering (ODR). ODR employs multi-buffering, dynamic rendering delay/acceleration, and input processing prioritization to reduce excessive rendering and ensure QoS satisfaction. ODR was evaluated in our private cloud and Google cloud. Evaluation results showed that ODR effectively removed excessive rendering, thus improving DRAM performance by 19% and reducing power usage by 16% over no FPS regulation. Better memory efficiency also allowed ODR to increase client FPS by 5.5%. Moreover, ODR reduced average MtP latency by more than 92% and outperformed existing FPS regulations. More importantly, ODR's high FPS and low latency make it feasible to deploy 3D applications to conventional public clouds.This work was partially supported by the National Science Foundation grants, 2221843, 2155096, 2215359, 2215193, and 2007718.Computer Scienc
PortraitEmotion3D: A Novel Dataset and 3D Emotion Estimation Method for Artistic Portraiture Analysis
No full textFacial Expression Recognition (FER) has been widely explored in realistic settings; however, its application to artistic portraiture presents unique challenges due to the stylistic interpretations of artists and the complex interplay of emotions conveyed by both the artist and the subject. This study addresses these challenges through three key contributions. First, we introduce the PortraitEmotion3D (PE3D) dataset, designed explicitly for FER tasks in artistic portraits. This dataset provides a robust foundation for advancing emotion recognition in visual art. Second, we propose an innovative 3D emotion estimation method that leverages three-dimensional labeling to capture the nuanced emotional spectrum depicted in artistic works. This approach surpasses traditional two-dimensional methods by enabling a more comprehensive understanding of the subtle and layered emotions often in artistic representations. Third, we enhance the feature learning phase by integrating a self-attention module, significantly improving facial feature representation and emotion recognition accuracy in artistic portraits. This advancement addresses this domain’s stylistic variations and complexity, setting a new benchmark for FER in artistic works. Evaluation of the PE3D dataset demonstrates our method’s high accuracy and robustness compared to existing state-of-the-art FER techniques. The integration of our module yields an average accuracy improvement of over 1% in recent FER systems. Additionally, combining our method with ESR-9 achieves a comparable accuracy of 88.3% on the FER+ dataset, demonstrating its generalizability to other FER benchmarks. This research deepens our understanding of emotional expression in art and facilitates potential applications in diverse fields, including human–computer interaction, security, healthcare diagnostics, and the entertainment industry.Electrical and Computer Engineerin
Alternative Treatments to Exercise for the Attenuation of Disuse-Induced Skeletal Muscle Atrophy in Rats
No full textThe prevalence of skeletal muscle atrophy, caused by disease and aging, is rising as life expectancy increases. Exercise is the most effective treatment option; however, it is often impractical for individuals suffering from disease or bedridden. The formulation of non-exercise-based interventions is necessary. This study assessed the impact of acupuncture (AC), electro-acupuncture (EA), and electrical stimulation (ES) on muscle mass and contractile properties in a model of casting-induced muscle atrophy. Sprague-Dawley rats (n = 40) were assigned to five groups: control (CON), cast (CT), cast receiving AC (CT-AC), cast receiving EA (CT-EA), and cast receiving ES (CT-ES) (n = 8 each). Treatments were 15 min and three times/week for 14 days. Contractile properties and protein markers of atrophy and inflammation were measured. Casting decreased muscle mass and fiber cross-sectional area, but AC, EA, and ES attenuated cast-induced muscle atrophy. All treatments increased peak twitch tension compared to CT. CT increased the protein levels of MAFbx and MuRF1, while AC, EA, and ES mitigated the elevation of these proteins. Our results indicate that acupuncture, electro-acupuncture, and electrical stimulation show promise as therapeutic strategies to counteract skeletal muscle loss and dysfunction resulting from disuse atrophy caused by injury, disease, and aging.Kinesiolog
Data-Centric Analysis of Security and Privacy of Containerized Applications
No full textContainerization has revolutionized software development and deployment by providing lightweight and portable encapsulations of applications and their dependencies. Technologies like Docker and Kubernetes have propelled the popularity of containers, offering a standardized and consistent environment throughout the software development lifecycle. Their agility enables swift deployment, scaling, and management of applications, fostering collaboration among diverse teams and accelerating development cycles. In the context of the burgeoning data generated by IoT devices, Edge devices are crucial for optimal performance. However, these devices face developmental challenges due to varying architectures. Containers mitigate these challenges by seamlessly deploying across diverse architectures, reducing development time. This pivotal role in the era of microservices architecture enables the modularization of applications into manageable components. Organizations increasingly adopt containerization for greater agility, resource optimization, and improved scalability, marking a paradigm shift in software development practices. The research aims to explore the vulnerability landscape of container applications on the Edge and traditional machines, evaluating different defense mechanisms and attacks. It identifies shortcomings in detection mechanisms, proposes improvements, and determines the most effective defense systems for specific types of attacks on containers. The ultimate objective is to promote awareness of container technology's potential for secure container deployment, monitoring, and prevention of security violations. The research seeks to securely deploy containers on diverse-edge architectures, including Raspberry Pi, SnapDragon, and Jetson Nano, defending against state-of-the-art attacks using advanced proposed defense systems that detect intrusions both statically and dynamically.Computer Scienc
Archaeological Report, No. 512
No full textThe Center for Archaeological Research (CAR) conducted an archaeological investigation of a section of the Upper Labor Acequia (ULA; 41BX2043) in Brackenridge Park, San Antonio, Texas. The work was in response to a request from the City of San Antonio (COSA) Public Works Department to provide information to the structural engineer for future remediation of the acequia. The project area is also within a previously designated site, 41BX1425 described as a large prehistoric campsite. The ULA is a contributing element to the Brackenridge Park National Register District and is also listed as a State Antiquities Landmark. The work described here was conducted under Texas Antiquities Permit (TAP) 31262, with Cynthia Munoz serving as the Principal Investigator.
The investigation, conducted in July 2023 and January 2024, included the excavation of three trenches to describe the exterior wall of the acequia and three test units to describe the interior wall of the acequia. In general, the acequia in the project area has been impacted, to varying degrees, as evidenced by highly fragmented wall sections, collapsing wall sections, missing stone, and/or wall sections with cracks. Trenching revealed that the exterior wall is comprised of stacked limestone rubble and larger stone that is not mortared. The interior of the wall consists of a mortared limestone facade. The foundation of the interior wall is limestone rubble resting on a clay matrix. No definitive acequia channel floor was found during the excavation. The test unit excavations within the acequia recovered modern debris, although a single flake was observed in the spoil pile of one of the test units.
All three external trenches contained prehistoric artifacts, including debitage and/or burned rock. The CAR documented an amorphous, ashy deposit in the floor of Trench 2 at approximately 80 cm below the top of the acequia wall. Designated Feature 1, it included burned earth, burned clay, ash, charcoal, and two burned rocks. The feature covers an area roughly 50 by 30 cm, with an unknown depth. A radiocarbon date on charcoal extracted from a sediment sample collected from the feature yielded a median date of 1306 cal BP with a two-sigma range of from 1345 to 1291 cal BP, documenting an occupation at the end of the Late Archaic Period. A burned rock cluster and a single flake was present in the western wall of the trench, roughly 30 cm above where the feature was defined. While the association of this material with Feature 1 is not clear, it is the case that an unknown portion of the feature was removed by the backhoe prior to the feature identification. Burned rock and debitage were also observed in Trench 1. Trench 3 contained these materials as well as faunal bone. Following the trench and test unit excavations the structural engineer, Shawn Franke, P.E., the City Archaeologist Matthew Elverson, and the THC’s Dr. Emily Dylla reviewed the exposed portion of the ULA.
Given the recovery during this investigation of cultural material and an archaeological feature with chronological information and integrity, CAR recommends that the deposits adjacent to the acequia on site 41BX1425 are eligible for listing on the National Register of Historic Places (NRHP) under criterion D, in that the deposits have yielded, and are likely to yield, information important to prehistory. In addition, the CAR recommends that this area of the site is eligible for designation as a State Antiquities Landmark (SAL) in that the site has the potential to contribute new and important information and thereby lead to a better understanding of Texas prehistory. Given that recommendation, and our current understanding of the proposed work, the CAR further recommends the development of a proactive testing program to mitigate impacts to this area of the site.
CAR will submit a site update (41BX1425) of this recent investigation to the Texas Archaeological Sites Atlas following comments by the COSA Office of Historic Preservation (OHP) and the Texas Historical Commission (THC), along with a GIS shapefile of the site and other information relevant to TAP No. 31262. Following completion of the fieldwork, all project-related materials, including the final report, were permanently stored at the CAR’s curation facility under accession number 2868.City of San Antoni
Molecular Mapping of In Situ CNS Lipids within EAE Tissue Using Matrix-Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry
No full textMultiple sclerosis (MS) is a neuroinflammatory disease that affects the central nervous system (CNS). A known pathological hallmark is lesions in the white matter of the brain and spinal cord. The lesions are a result of demyelination, the breakdown of myelin, which is a protective sheath made up of proteins and fatty substances. Myelin wraps around the nerve axons and allows for smooth electrical impulses throughout the CNS. Demyelinated axons have exposed fiber that disrupts nerve communication. The mechanism for demyelination has yet to be determined and the role that lipids play in the demyelination mechanism requires deeper exploration.
In the first study presented in this thesis, trapped ion mobility mass spectrometry time of flight (TIMS-TOF) with a matrix-assisted laser desorption/ionization (MALDI) source is used to create molecular mass maps of in situ lipids from experimental autoimmune encephalomyelitis (EAE) naïve mouse brain tissue. Matrices 2,5-dihydroxybenzoic acid (DHB), 1,5-diaminonaphthalene (DAN), and 9-aminoacridine (9-AA) are chosen to determine which lipids each matrix is successful in ionizing in both positive and negative mode. The mass spectrometry imaging (MSI) for mouse brain tissue shows that DHB is more successful in ionizing lipids in the positive mode. While DAN and 9-AA are more successful in the ionization of lipids in the negative mode. This could be because DHB is an acid that will readily give up a proton, and DAN and 9-AA are basic thus promoting deprotonation. The mass spectra show that DHB can ionize lipids of higher molecular weights, while DAN and 9-AA work better for lower molecular weight lipids. DHB and DAN were shown to be successful in mapping phospholipids, while 9-AA was the most successful in mapping sulfatides.
In the second study mouse spinal cord tissue was coated with matrix 9-AA. MSI was performed in negative mode ionization using MALDI-TOF. The MSI results show that the method used to map lipids in mouse brain tissue is unsuccessful for spinal cord tissue. MSI showed no localization of lipids on the tissue due to the analyte being washed to the edge of the tissue. Yet the mass spectra generated can be used to determine which lipids are present in the sample. Future work requires optimization of spinal cord tissue sample preparation to enhance ion signal localization on the tissue. Hematoxylin and eosin staining were used in both studies to confirm tissue histology. The staining was successful in confirming the histology of the tissue after collecting MSI data. The findings from both studies will be used to further investigate lipids in MS. Further analysis is required to confirm the role they play in the demyelination mechanism.Chemistr
A Novel Breast Ultrasound Image Augmentation Method Using Advanced Neural Style Transfer: An Efficient and Explainable Approach
No full textClinical diagnosis of breast malignancy (BM) still remains challenging, with deep learning (DL) models showing promise for early detection. However, their performance is often hindered by overfitting due to limited breast ultrasound (BUS) image data and privacy concerns associated with datasets. Image augmentation is crucial to improve DL model performance, yet existing approaches are often opaque and computationally intensive. This thesis introduces a novel augmentation method for BUS images, integrating advanced neural style transfer (NST) and Explainable AI (XAI) on a GPU-based parallel infrastructure. Using the Horovod framework on an 8-GPU DGX cluster, we achieve a 7.71x speedup while preserving accuracy. By evaluating our proposed model on 800 BUS images (348 benign, 452 malignant), it demonstrates a 37.26% improvement
in classification accuracy when comparing pre-augmented and post-augmented data using a fully fine-tuned ResNet50 model.Computer Scienc
Combined Emission Economic Dispatch using Quantum-inspired Particle Swarm Optimization and its Variants
No full textThe ever-increasing electricity demand, its dependency on fossil fuels, and the consequent environmental degradation are major concerns of this era. The worldwide domination of fossil fuels in bulk electricity generation is rapidly increasing the emissions of CO2 and other environmentally dangerous gases that are contributing to climate change. The economic and emission dispatch are two important problems in thermal power generation whose combination produces a complex highly constrained nonlinear optimization problem known as combined economic and emission dispatch. The optimization of combined economic and emission dispatch aims to allocate the generation of committed units to minimize fuel cost and emissions, simultaneously while honoring all equality and inequality constraints. Therefore, in this article, we investigate a solution of the combined economic and emission dispatch problem using quantum particle swarm optimization and its two modified versions, that is, enhanced quantum particle swarm optimization and quantum particle swarm optimization integrated with weighted mean personal best and adaptive local attractor. The enhanced quantum particle swarm optimization algorithm achieves particles’ diversification at early stages and shows good performance in local search at later stages. The quantum particle swarm optimization integrated with weighted mean personal best and adaptive local attractor boosts search performance of quantum particle swarm optimization and attains better global optimality. The suggested methods are employed to achieve solution for the combined economic and emission dispatch in four distinct systems, encompassing two scenarios with 6 units each, one with a 10-unit configuration, and another with an 11-unit setup. A comparative analysis with methodologies documented in existing literature reveals that the proposed approach outperforms others, demonstrating superior computational performance and robust efficiency.Electrical and Computer Engineerin
The Curriculum in IDD Healthcare (CIDDH) eLearn Course: Evidence of Continued Effectiveness Using the Streamlined Evaluation and Analysis Method (SEAM)
No full textMedical professionals are rarely trained to treat the unique healthcare needs and health disparities of people with intellectual and developmental disabilities (IDD). The Curriculum in IDD Healthcare (CIDDH) eLearn course aims to redress gaps in the delivery of medical care to people with IDD. An initial comprehensive evaluation of CIDDH in-person training content had previously underscored its knowledge and skill transfer efficacy for Mississippi healthcare providers. Training content has recently become available to medical professionals nationwide through an online self-paced modality to address physicians' IDD education needs. This study introduces and applies a new evaluation framework called SEAM (Streamlined Evaluation and Analysis Method) that offers a promising avenue for rendering a follow-up appraisal after rigorous evidence of program effectiveness has been previously established. SEAM reduces the data-reporting burden on trainees and maximizes instructor–trainee contact time by relying on an abbreviated post-only questionnaire focused on subjective trainee appraisals. It further reduces methodological and analytical complexity to enhance programmatic self-assessment and facilitate sound data interpretation when an external evaluator is unavailable. Ratings from a small sample of early-cohort trainees provide an important test of effectiveness during CIDDH's transition to online learning for clinicians nationwide. Using SEAM, CIDDH achieved high ratings from this initial wave of trainees across various evaluative domains. The study concludes by highlighting several promising implications for CIDDH and SEAM.Sociology and Demograph
Two Essays on the Impact of Marketing Strategy on Financial Market Metrics Using Natural Language Processing
No full textThis dissertation consists of two essays examining the impact of critical aspects of marketing strategy on financial outcomes such as risk, returns and firm value.
Essay 1 examines how a firm's strategy-scope states change over time and identifies the financial consequences of the changes. First, using a grounded-theory framework, a measure of firm-strategy scope is developed using a dataset of 29,340,577 sentences from 60,280 10-K reports of 4,540 B-to-B companies spanning over 21 years. Second, a Hidden Markov Model shows three latent strategy scope states – traditional, relational and diffused that B-to-B firms occupy at any given time with 36% of firms switching their state of strategy scope at least once and 20% switching two or more times over the 21-year period. Third, the article empirically identifies the consequences of changing strategy-scope states. Changing strategy scope poses a tradeoff between lower risk and lower returns for executives. Results show that every time a firm switches its strategy scope state, it reduces idiosyncratic risk by .40 standard deviation, and systematic risk by .21 standard deviation, while decreasing abnormal returns by 6.5%. We find that systematic risk is minimized when transitioning from diffused to relational strategy states but peaks when transitioning from diffused to traditional strategy states. Conversely, idiosyncratic risk is higher when transitioning from relational to diffused state but reaches a low point when transitioning from traditional to diffused state. Abnormal returns are maximized when transitioning from traditional to relational state. Furthermore, services companies experience lower idiosyncratic and systematic risk with higher returns.
Essay 2 addresses the evolving landscape of environmental, social, and governance (ESG) considerations and their impact on firm value. Employing a deep learning model on earnings call transcripts, and using the distinction between material and nonmaterial ESG factors, this research assess their influence on firm value. Contrary to prevailing expectations, an emphasis on nonmaterial ESG factors is found to detrimentally affect firm value, particularly in regulated industries. Conversely, material ESG emphasis yields positive effects, albeit to a lesser extent. The study underscores the importance of strategic ESG integration and offers a practical tool for executives and investors to evaluate and compare companies' ESG strategies and their impact on firm value.
Together, these essays contribute to a nuanced understanding of strategic decision-making and its financial ramifications in an increasingly complex and socially conscious business environment.Marketin