Treasures @ UT Dallas
Not a member yet
7697 research outputs found
Sort by
Efficient and Field Deployable Personalization of Prescriptive Amplifications in Hearing Aids
No full textStandard hearing aid prescriptions typically use audiograms to set amplification levels across a
number of frequency bands based on averages of populations with similar audiograms. In previous
studies, it has been established that personalization of prescriptive amplification in hearing aids
improves hearing perception. This dissertation addresses achieving personalization of
amplification in hearing aids by developing a training-efficient and field-deployable approach.
This personalization approach is developed based on Bayesian machine learning within a multi-
band framework. The training efficiency aspect of the developed personalization approach is
achieved via paired comparisons and its field deployable aspect is achieved via implementation of
smartphone apps to run it in real-time on smartphone platforms. These attributes enable the
personalization of amplification to be carried out in real-world audio environments in an easy-to-
use manner. Subject testings involving speech in noisy background have indicated that the
personalized amplification settings by the developed personalization approach are preferred over
the standard prescriptive amplification settings. Two hearing aid amplification prescriptions are
examined to show that the developed personalization approach is general purpose in the sense that
it can be applied to any prescription
Examining the Impact of Relationship Education on the Dating Behaviors and Personal Wellbeing of Single Emerging Adults
No full textRomantic relationships have a significant impact on various aspects of one’s health. For many
young adults, entering and maintaining long-term romantic relationships is a key marker of
adulthood, and the quality of these relationships plays a central role in mental health outcomes.
Therefore, it is imperative that young adults are equipped with the appropriate skills to promote
healthy relationships. This study evaluated how romantic competence training – a skills-based
model of adaptive relationship functioning -- impacts single emerging adults’ dating behaviors,
mental health, and personal wellbeing outcomes. To address these aims, this study used a pre-
post-post experimental design to administer a 2-week relationship skills training (total of two 2-
hour sessions) to determine the effects of romantic competence training on the dating behaviors
of single emerging adults and downstream mental health and wellness outcomes over four
months. Results from this study replicate previous study findings that demonstrate the efficacy of
romantic competence training on improved relationship knowledge, such as an increased ability
to manage and cope with romantic relationships and improved skills like insight. Although these
results show potential for improving relationship knowledge and skills, future studies can
examine if increased knowledge leads to adaptive behavioral changes. Findings from this study
can be used to underscore the importance of integrating relationship education into schools, as
this may ensure that young adults are equipped with the necessary skills to navigate dating and
romantic relationships effectively
Porous Polyphenylene Polymers
No full textCovalent organic frameworks (COFs) have been an upcoming field of porous polymers since the
early 2000s. They are entirely composed of lightweight elements (C, N, O, B, H, etc.) and held
together by covalent bonds. Some of their attractive physical properties includes their crystallinity,
low-density, and permanent pores. The construction of COFs begins with the synthesis of building
blocks, which are held together by covalent linkages such as boron ester, imine, azine, etc. These
building blocks are then brought together by solvothermal synthesis, a common method for COFs.
Typically, the monomers are added to compatible solvents with an acid catalyst in a glass ampoule,
undergoes degassing via freezer-pump-thaw cycles, flamed sealed and placed in a designated oven
for a few days. Initially, the kinetic amorphous intermediates dominate and then the amorphous
solid undergoes error correction over time to achieve the favored thermodynamic product. Error
correction is possible due to the dynamic covalent chemistry observed between the compatible
monomers undergoing condensation reactions. This leads to COF layers forming that are held
together by noncovalent interactions such as − interactions, hydrogen bonding, and the
alignment of dipole moments. Therefore, the crystallinity, permanent porosity, and low-density of
COFs can be attributed to dynamic covalent chemistry, being held together by covalent bonds, and
reinforced by non-covalent interactions, and being comprised of lightweight elements. In addition,
COFs are tunable with a library of monomers to choose from, by modifying the monomers before
COF synthesis or modifying the COF post-synthesis. Thus, the goal of this research is to leverage
the tunability of COFs with various monomers to elucidate design protocols for the purpose of
synthesizing stable mesoporous COFs and broadening the understanding of what is required to
achieve them.
The first chapter describes COF design and applications with scientific literature. In the second
chapter, a novel monomer was synthesized with the intended application for the synthesis of a new,
mesoporous 2D-COF. This monomer, 1, features five phenyl rings in a row contributing length to
attempt the synthesis of a new mesoporous COF. It was primarily designed to be explored with
the novel amide monomer, synthesized in the Smaldone lab, to determine if the torsion of the
phenyl rings can be overcome by the interlayer hydrogen bonding of the amides. 1 was further
explored with other monomers and resulted in two new POPs
The Disordered Spirit: A Portrait of Francisco Amighetti as Seen by Laura Goldstein
No full text“The Disordered Spirit: Francisco Amighetti as seen by Laura Goldstein” addresses the paradox
of the fragment and the whole inherent in the ideas of “essence” or “truth” through the work of
Costa Rican artist and poet Francisco Amighetti, and offers some alternative perspectives to the
anxiety of loss in literary translation. This investigation emphasizes the overuse of discussions on
loss in translation and argues for the value of the fragment, particularly in Amighetti’s work,
which falls in the Modernist period when poets and artists turned to the fragment through style,
technique, and a confrontation with the past, but also more broadly to argue that the fragment has
value through the choices of creative processes, language, and expression, the multiplicity of
subjective experiences, order and disorder, and through the nature of memory and our universe.
The dissertation also analyzes the creative work of Romanian-Brazilian writer Ștefan Baciu, who
wrote poems responding to fragments of Amighetti’s poems, letters, and prose, and finally
includes creative work by the author of the dissertation in the form of original poetry, poetry in
translation, visual art (prints) and memoir, proposing that a translator can reveal the multiplicity
of subjective experiences through the inclusion of their original creative work, especially when
the translated poet is excluded from the canon as Francisco Amighetti and other Costa Rican
poets have been
The Enchantment of Embodied Living: Aesthetic Engagement and Modernity in Contemporary British Fiction
No full textThis dissertation analyzes a set of contemporary British novels that challenge a pervasive trend
in Western philosophy, namely Cartesian dualistic thinking. Cartesian philosophy gives supreme
importance to the human mind because of its association with reason and rationality and
denigrates the human body. The analysis in the dissertation builds on recent work by literary
scholars such as Elizabeth S. Anker, who maintains that the phenomenological understanding of
the human body can restore faith in corporeality and correct certain shortcomings of instrumental
reason. She adds that literature, because of its corporeal and embodied nature, has the potential to
envision a different conception of being human. However, there is no scholarship in
contemporary British fiction that examines how embodiment or corporeality works as an antidote
to instrumental reason. The dissertation argues that the selected novels create an alternative
imaginary of modernity exemplified by embodied living. The select group of novels in this
dissertation are Home Fire (2017), Exit West (2017), Happiness (2018), and Spring (2019). This
dissertation shows that these works of contemporary British fiction rehabilitate aspects of liberal
modernity and rationality that they represent as redeemable. These novels link a rational
understanding of political reality to embodied experience. For example, Home Fire (2017) and
Exit West (2017) develop both in their aesthetic forms and character development a more
expansive and liberating notion of autonomy, progress, and freedom by reorienting the idea of
agency and personhood around embodied and corporeal experience. Happiness (2018) articulates
its vision of an intertwined existence by challenging the anthropocentric worldview and
competitive market economy. Spring (2019) advances its aesthetic theory of embodied art to
counteract the onslaught of neoliberal commodity art
Modernizing Hardware Circuits Through High-level Synthesis
No full textRegister-Transfer Level (RTL) languages, such as Verilog HDL, have been fundamental in
digital hardware design for decades, resulting in the creation of numerous legacy hardware
designs. However, due to rapid technological advancements, expanding application domains,
and evolving design requirements, many of these legacy designs are becoming outdated.
Adapting these designs to newer technology platforms or meeting modern design requirements
typically involves laborious manual re-scheduling and re-verification in Verilog, which is a time-
consuming and inefficient process. In contrast, High-Level Synthesis (HLS) offers significant
advantages over traditional handwritten RTL. HLS allows for the generation of diverse
hardware architectures from the same behavioral description to address different operational
demands, such as varying frequencies or implementation platforms like ASICs/FPGAs. An
RTL-to-C compiler serves as a bridge between RTL and high-level design. It enables the
legacy RTL designs to be automatically re-synthesized for a different technology platform or
different design requirements through High-Level Synthesis. This is called Modernization.
This dissertation addresses the modernization of legacy hardware by introducing an efficient
RTL-to-C compiler optimized for HLS.
A key feature of HLS is Design Space Exploration (DSE), where designers can control
synthesis attributes like loops and arrays using synthesis directives or pragmas. By adjusting
these pragmas, a single behavioral description can produce multiple microarchitectures
corresponding to different quality metrics such as area, performance, and power. However,
exploring this vast design space comprehensively is challenging due to the large number of
possible pragma settings. To efficiently find the Pareto-optimal designs, heuristic algorithms
are employed. However, multi-threaded execution of these algorithms requires multiple
licenses. ASIC HLS tools typically incur high licensing costs, whereas FPGA HLS tools offer
free licenses. This dissertation proposes a smart technique to leverage the free FPGA HLS
tool to run the DSE heuristic algorithm in multiple threads and use the results to explore
the design space for the ASIC platform. Additionally, this dissertation introduces security as
a novel dimension in design space exploration, alongside traditional quality metrics. This
enhancement aims to modernize legacy RTL designs to withstand potential side-channel
attacks.
Another interesting finding about HLS tools is that the choice of input behavioral language
significantly impacts Quality of Results (QoR) due to the variations in parsing mechanisms
of different language parsers. To optimize this process, this dissertation presents a novel
approach based on Graph Convolutional Networks (GCNs) to predict and automatically
translate behavioral descriptions into the most suitable language, thereby enhancing design
optimization
Ordinal Patterns-based Time Series Analysis
No full textMost real-world phenomena exhibit non-stationary behavior, where the statistical properties
of the underlying process change over time. Most pre-existing techniques perform very
well for time series realized from stationary processes, but fail for non-stationary processes.
Traditional stationary techniques may not adequately capture the dynamics of the data;
neglecting non-stationarity can lead to erroneous conclusions and flawed models.
In this dissertation, we introduce a novel technique to generate surrogate data for time series
measured from non-stationary systems. The surrogates generated are called Order Preserving
surrogates and are defined in a way that preserves the ordinal patterns of the original signal up
to a predefined length. Recently, there has been growing interest in studying non-stationarty
time series using ordinal partition transition networks (OPTN) generated from them. Our
surrogate method preserves the OPTN generated from the original signal, such that the
OPTN will be the same for all the surrogates of the same signal.
We have applied our novel approach for generating surrogates to two separate projects.
Our first project focuses on detecting nonlinearity in possibly non-stationary signals using
numerous discriminating statistics. Our second project uses the Order Preserving surrogates
to detect spatial patterns between two signals evolving over time. We use the Order Preserving
surrogates in combination with wavelet coherence to detect statistically significant correlation
between signals
Cataclysm, Calamity, and Catastrophe: Refining Craft, Speculating Futures, and Metaphorizing Truths Through Miniatures
No full textThis MFA Thesis delineates the story, the afflatus, and the making of the work exhibited in
Cataclysm, Calamity, and Catastrophe. The events that turned an Engineering student into an
Art student, the artists that inspired and drove the work to new horizons, and the refining of craft
are all equally influential in the development of a miniature sculptural practice shown in the
following 3 series. Metaphorical Monsters, Physical Phobias, and Forefathers Fossils reduce
perceived and wholly imagined futures of eschatological events into miniatures rendered in
complex detail which are intended to be viewed with the same level of intimacy and attentive
perception as was required to craft them.
This pursuit of craft and condensing a story into a single sculpted moment in my work aims to
render “frames of finality” that allow the viewer to contend with both fantastical and realistic
scenes of the end of civilization and how that theme spurned my practice of conservation
through making
Ultra Low-Power Micro Electro-Mechanical Sensors as Neural Computing Units
No full textThis thesis investigates the design, fabrication, and operation of ultra-low-power MEMS devices
as neural computing units and integrated sensors, advancing MEMS technology for hardware-level
data processing and decision-making. The research addresses key challenges in MEMS design,
including reducing power consumption in digital accelerometers and implementing coupled
MEMS neurons capable of real-time signal classification without external digital processing.
A novel, digitally operated MEMS accelerometer with 8-bit resolution was introduced, utilizing
electrostatic tuning to achieve high-precision measurements while significantly lowering power
requirements. By eliminating the analog front-end, this accelerometer achieves substantial
efficiency improvements, making it ideal for energy-constrained applications.
Further, this work presents a groundbreaking approach to neuromorphic computing using
mechanically and electrostatically coupled MEMS devices. These MEMS neurons mimic neural
network functionality, leveraging mechanical dynamics to classify signals and process information
autonomously. This innovation transforms MEMS from passive sensors into active computing
units, enabling real-time computation with minimal power consumption.
The thesis also addresses critical fabrication challenges, particularly stiction, a significant obstacle
in MEMS reliability. A novel anti-stiction method, the naphthalene-IPA sublimation technique,
significantly improved release yields for delicate microstructures, enhancing device reliability and
broadening the scope of MEMS applications.
In summary, this research expands the frontiers of MEMS technology, paving the way for highly
efficient, autonomous sensor and computing units suited for the Internet of Things, wearable
devices, and other power-sensitive environments. Through innovations in MEMS design and
fabrication, this thesis demonstrates the transformative potential of MEMS as integrated sensing
and computational platforms
Systematic Methods to Analyze and Recognize Illicit Information Manipulation
No full textImprovements in AI-synthesized content present challenges and opportunities for improving
the quality and integrity of online information sources. Users increasingly depend on online
information sources such as search engines or photos. This work examines three applications
for AI synthesis to secure online information sources. First, this article will describe how
to use AI text synthesis techniques to identify existing malicious search engine poisoning
attacks. Second, this work will present how to identify AI-synthesized images based on
derived model signatures. Finally, this dissertation will demonstrate how users perceive
current AI-synthesized images and how to leverage these insights to develop more robust
detection models for this content. As AI-synthesized content tools and illicit information
manipulation continue to grow, they will impact positively and negatively impact online
information sources. As a result, developing defenses and ways to leverage these tools is
critical to protect online information sources.
Thesis: Illicit information manipulation can be systematically recognized by analyzing fundamental characteristics