137091 research outputs found
Sort by
Interactive Sonification of 2D Quantum Systems
No full textPresented at the 30th International Conference on Auditory Display (ICAD 2025)This paper presents novel sonification methods for the auditory representation of 2D quantum systems and their temporal evolution, encompassing examples such as the tunnel effect, single and double slit interference or behavior of light or matter in a given potential function. The simulation involves numeric integration of the Schrödinger equation. The sonification is based on three paradigms: (a) scanning the probability amplitude along a 1Dmanifold (i.e. curve) as a waveform, (b) probing the probability amplitude as spectral activation along a 1D-manifold, (c) traversing the full 2D field as an audification. We illustrate the methods with sonification examples, discuss what can be learned about the behavior of quantum systems in non-stationary transitions and propose application scenarios in physical, musical and educational contexts
Auditory Graphs for Stochastic Processes: A Case Study on Mathematical Accessibility
No full textPresented at the 30th International Conference on Auditory Display (ICAD 2025)The demand for data scientists has been rapidly increasing in recent years. However, visually impaired individuals face significant challenges in interpreting stochastic process data, which are frequently used in various fields such as finance. Addressing these barriers through auditory graph techniques could help expand career opportunities for visually impaired individuals. In this study, we evaluate whether auditory graphs, which map the values of functions to variations in audio pitch, enable five visually impaired students from grades 5, 9, and 12 to differentiate between graphs of Brownian motion. Experimental results suggest that auditory graphs effectively convey the distinguishing features of Brownian motion, indicating their potential for supporting mathematical accessibility in stochastic contexts. Due to the small number of participants, this study should be regarded as a pilot study, and further research with larger sample sizes is required to confirm generalizability
Metabolomic Analysis of Chemical Interactions between Burkholderia thailandensis and Host to Inform Alternative Treatments
No full textBurkholderia pseudomallei, a soil-dwelling pathogen, is the etiological agent for the human disease melioidosis and classified as a Tier 1 Select Agent, posing a serious global threat. Due to the absence of an FDA-approved vaccine against melioidosis as well as the multi-drug resistance of B. pseudomallei, there continues to be high-fatalities from this infection in the tropics and sub-tropics. The lethality of B. pseudomallei is further amplified by its transmissibility through inhalation, ingestion, and skin abrasion as well as its diverse clinical presentations. Production of virulence factors and modification of host cell functions triggered by Burkholderia infection constitute important mechanisms in the pathogenesis of this respiratory infection, including evasion of host immune responses. Currently, the metabolic responses during host-pathogen interactions that could inform medical countermeasures are not well-understood, but recent advances in mass spectrometry methods are enabling the narrowing of this knowledge gap. Towards this end, we coupled untargeted metabolomic analyses of Burkholderia models of infection (Burkholderia thailandensis, a validated surrogate) in mammalian hosts with high-throughput screening methods of candidate probiotics in order to inform alternative treatment approaches to combat Burkholderia infection. Through dual metabolome profiling, we characterized the pathogen’s chemical arsenal employed during co-culture with mammalian cells where it produces proposed virulence factors such as burkholdacs, bactobolins, acybolins, acyl-homoserine lactones, and 4-hydroxy-3-methyl-2-alkylquinoline. Conversely, we robustly captured the chemical signature of the host’s immune response to the presence of this pathogen, revealing the activation of several immune lipid pathways including the release of prostaglandins, as well as major disruptions to the host’s primary metabolic pathways such as the tricarboxylic acid cycle upon exposure to Burkholderia. Our characterization of the host-pathogen chemical interactions enabled further investigation into candidate probiotic bacteria as a potential medical countermeasure for Burkholderia infection. Using whole-cell matrix-assisted laser desorption/ionization (MALDI) mass spectrometry to screen candidate bacteria against Burkholderia to further prioritize strains for in vivo protection assays, we identified a Bacillus velezensis CP8 strain that elicits natural product production by the model pathogen and results in growth inhibition. We also establish the potential application of a Brevibacillus borstelensis CP19 strain as an airway probiotic treatment against B. thailandensis infection models in vivo. Thus, we present the use of various mass spectrometry-based tools as an effective method for studying host-pathogen interactions and screening airway probiotic candidates as alternative treatment methods for antibiotic resistant respiratory pathogens like B. pseudomallei.Ph.D.Chemistry and Biochemistr
Preliminary Analysis of Northwest Corridor Revenue from General Purpose and Managed Lanes by Household Income Group
No full textThis master’s thesis research presents Georgia Department of Transportation (GDOT) expenses to construct reversible managed lanes along the Northwest Corridor (NWC), takes a first-cut at estimating costs to operate the system and the time savings benefits of the system, and allocates benefits and costs across user groups by household income, using demographic results from a 2022 study. This research consisted of calculating corridor improvement costs and revenue streams that agencies face during the construction of managed lanes, beginning with construction costs (without maintenance). Gas taxes and toll expenditures were also estimated from traffic volumes and fuel consumption modeling. The Amount of gas tax spent for each household income group was then distributed among all vehicles ages to generate how much each income group is spending on gas tax alone each year, using SRTA studies. Annual corridor toll revenues were also estimated using public data. These costs were then allocated to income groups using data from a previous demographic study of corridor users. Payback periods were also analyzed to see how long it would take to completely pay off the corridors, assuming no major increase in vehicle activity, tolls, or average vehicle fuel economy.M.S.Civil Engineering/City and Regional Plannin
There's an Archivist for That?! Chapter 2
No full textInterview portion of Lost in the Stacks, episode 648. Features interview with Allison Schein, Director of Archives and Rights Management for The Atlanta Journal-Constitution. Allison discusses how she got into the world of archives from her background as an audio engineer, and what she does in her current role at the AJC.Interview portion of Lost in the Stacks, episode 648. Features interview with Allison Schein, Director of Archives and Rights Management for The Atlanta Journal-Constitution. Allison discusses how she got into the world of archives from her background as an audio engineer, and what she does in her current role at the AJC
Broad-coverage Multi-cultural Stance Classification
No full textWe present Stanceosaurus, a new corpus of 28,033 tweets in English, Hindi, and Arabic annotated with stance towards 251 misinformation claims. As far as we are aware, it is the largest corpus annotated with stance towards misinformation claims. The claims in Stanceosaurus originate from 15 fact-checking sources that cover diverse geographical regions and cultures. Unlike existing stance datasets, we introduce a more fine-grained 5-class labeling strategy with additional subcategories to distinguish implicit stance. Pre-trained transformer-based stance classifiers
that are fine-tuned on our corpus show good generalization on unseen claims and regional claims from countries outside the training data. Cross-lingual experiments demonstrate Stanceosaurus’ capability of training multi-lingual models, achieving 53.1 F1 on Hindi and 50.4 F1 on Arabic without any target-language fine-tuning. Finally, we show how a domain adaptation method can be used to
improve performance on Stanceosaurus using
additional RumourEval-2019 data. We make Stanceosaurus publicly available to the research community and hope it will encourage further work on misinformation identification across languages and cultures.UndergraduateComputer Scienc
Data and Computation-efficient Deep Learning for Multi-agent Systems
No full textThe primary goal of this research is to build data and computation-efficient deep learning methods for multi-agent systems. Multi-agent systems are present in a wide range of domains, from physical systems (e.g., molecules, planets) and biological systems (e.g., host-pathogen interactions, neurons) to social systems (e.g., covid-19 spread, games with human players). Although these systems have significant real-world applications, mathematical modeling of their often unknown dynamics is challenging. Deep learning offers a data-driven approach to modeling these systems without requiring extensive domain knowledge. However, collecting sufficient training data is difficult, as these systems evolve over time, and we may not even detect when the underlying dynamics change. Moreover, multi-agent systems are often driven by a large number of agents, making learning and prediction computationally expensive and inefficient. This thesis explores these challenges by developing innovative algorithms and neural network designs that can efficiently learn representations of the spatial arrangement of agents, forecast their trajectories and state transitions, and uncover hidden interaction graphs in unstructured and structured multi-agent systems, considering data and computation constraints.Ph.D.Electrical and Computer Engineerin
Coupled Auto-Enrollment and Speaker Identification Platform for Real-Time Applications
No full textWhile many individuals naturally experience a natural cognitive decline due to old age, a significant number of individuals experience a faster cognitive decline, which may manifest itself as dementia. While there are many studies on preventative care for those individuals, the quantity of social interactions for those individuals, on a daily basis, plays a large role in whether or not they experience a rapid cognitive decline.
A monitoring system capable of identifying speakers with very little data would a crucial first step to understand the nature of those interactions, but it is challenging to identify speakers with close to zero training data. There have been many advancements in speaker identification in the field of speech processing. Speaker identification is a different type of classification problem, given that it requires an enrollment component to it. Some high performant frameworks may sometimes need long durations of audio for each speaker, and while there has been progress on diminishing the amount of training data to develop such speaker identification systems, research on these topics are nevertheless important. A system capable of quickly enrolling speakers for identification could lead to many more applications beyond preventative health care.
One example could be that movies, TV shows and other forms of media could have enhanced subtitles. While present day subtitles inform the audience what is being said, it doesn't always inform who is speaking. Hard-of-hearing audiences suffer from such a visualization, as they can only infer that the person speaking is present on the screen. If such a framework is capable of running as an online algorithm, thousands of hours of videos/podcasts could be properly tagged to assist deaf individuals. Additionally, if such a system was capable of running in real-time, journalistic interviews, presidential debates, sports commentaries could also benefit from such an expansion.
This work builds the backbone and a functional system capable of doing speaker identification in real-time, aiming to bridge the gaps for the purposes of monitoring the quantity of interactions for at-risk populations. It paves a different pathway for an individual to map the interior of a space (e.g. home, office), determine optimal locations to place microphone arrays, set up a server and edge nodes, and run the aforementioned autonomous system capable of detecting new classes (i.e. speakers) with only 2.5 seconds of audio, auto-enrolling new speakers, and re-identifying the speakers in real-time.Ph.D.Machine Learnin