University of Texas at Dallas

NSF Awards · FY 2024 · 2024-10

Artificial intelligence (AI) and machine learning (ML) has been widely and successfully used in many fields including transportation, autonomous driving, chip design, etc. Considering the profound impact of AI as a potent force of transformation across various societal domains, AI ethics has garnered significant scrutiny. AI systems trained on biased data can perpetuate or amplify negative biases, with profound implications for areas like criminal justice, hiring, and lending, where biased AI could lead to unfair or discriminatory outcomes. Designing an ethical AI system has significant social and political value. As AI models grow, the demand for cyberinfrastructure (CI) support becomes substantial. Much research has focused on designing high-performance computing (HPC) infrastructures to accelerate AI/ML. However, support from CI for ethical AI is lacking, primarily due to distinctive constraints introduced by ethical considerations. Notably, such ethical constraints or objectives integrated with AI algorithms can slow down the inference and training processes. Conversely, without consideration of ethical AI, traditional CI technologies such as quantization and approximation might compromise AI ethics, even if they expedite the computation. This project will establish interactive and integrated training for building high-performance ethical AI with three interdisciplinary training programs across philosophy, ethical AI, and HPC. These include nine training modules and activities for sustainability and fostering community. The goal is to fill the gap between CI and ethical AI and AI ethics and train both CI contributors and CI users to build high-performance ethical AI. The training programs include: 1) Philosophical AI ethics training for CI contributors and ethical AI designers; 2) Ethical AI training for CI contributors; 3) CI software and hardware technologies training for ethical AI designers. Moreover, several hands-on projects are proposed to deepen trainees’ understanding of those programs, including hardware acceleration for machine learning models, ethical AI implementation, etc. The long-term goal is to boost the adoption of new "Computing+AI+Ethics" to multidisciplinary students and researchers from different STEM domains. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-10

While integration of multimodal information has been widely researched, the difficulty in carrying out this integration in a domain-agnostic, generic manner has been problematic. Past results have emphasized the need for addressing multimodal integration in a domain-specific manner. For instance, cardiac rehabilitation is a diverse range of practices for restoring individual's functioning. Unfortunately, only 30% - 40% of patients report regular exercise at six months after discharge and 39%-45% of these patients suffer from at least one readmission within one year. These poor outcomes motivate the need for using technology for remote monitoring in this multi-modal system. The overarching goal of this proposal is to understand how multiple cardiac experts view multimodal data, decide on the progress, and the variations/biases among the experts' views and decisions. This understanding would help us design and build a multimodal visualization and progress-tracking system that can provide meaningful information to stakeholders. Though the proposed algorithms for multimodal data analysis are tightly tied to the biomedical domain, the derived knowledge and understanding would be useful to other domains using multimodal sensing. Results, metrics, and algorithms from this research will be published widely in high-quality academic journals and conference proceedings. Integrating, visualizing and tracking progress using multimodal data is highly domain dependent. In this project, cardiac tele-rehabilitation deployed in-home is the target domain, generating multimodal data, with their diverse data characteristics and varied timeframes. Research challenges in domain-specific multimodal integration typically include the characteristics of the multimodal data as well as the domain-specific needs. For multimodal integration for cardiac rehabilitation, the challenges include: (i) Integrating the multimodal data with diverse types of data with varying temporal characteristics to find relationships among potential adverse events; and, (ii) Possibilities for using mobile and wearable sensors to provide opportunities for personalization both in the rehabilitation and in the multimodal data integration. The proposed system, in the form of a mobile app, will democratize data acquisition. This, in turn, could lead to a better understanding of bias among experts and possible strategies for mitigating the bias and provide appropriate feedback and nudges to patients. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-10

Currently, up to 60% of consumed energy is related to electrical energy which undergoes a large amount of wastage in power electronics systems, and by 2030, it is expected that 80% of all electric power will flow through power electronics systems. Key to this wastage is the limited efficiency and scalability of power devices that are the building blocks of power conversion systems. For the projected compact high-power electronics, ultra-wide bandgap (UWBG) semiconductors, such as beta-phase of gallium oxide (β-Ga2O3), can be a viable candidate to enable significantly efficient, smaller, and faster power switches, replacing silicon. This collaborative project aims to develop a scientific base and engineering for robust multi-kV ampere class compact packaged UWBG β-Ga2O3 power devices by addressing existing challenges at both the device level and the packaging level. The integrated education plans will train the next generation of UWBG engineers and researchers to maintain the competitive vitality of the U.S. power electronics workforce in light of the trend towards high voltages, high power densities, and high temperatures. The project will also involve incoming first-year undergraduate students in this research program through the Clark Summer Research Program at the University of Texas at Dallas and the First Year Honors Mentor Program at Iowa State University. This project will establish a comprehensive understanding of β-Ga2O3 devices under extreme fields, high temperatures, and defects, and perform a holistic improvement integrating material properties, device design, and packaging techniques. At the device level, the objectives are (i) robust high field and thermal management combining (ultra)high permittivity and high thermal conductivity dielectric, guard rings, and substrate thinning, (ii) Schottky barrier engineering with optimized contact configuration to enable high surface breakdown field, improved thermal stability, and low loss, and (iii) defect mitigation in large-area power devices through identifying defects in drift layers and interfaces, and coordinated process development. At the packaging level, this pioneering proposal will investigate and integrate two novel insulation materials and systems: (I) two particular high-temperature liquid dielectrics as pure and nanodielectric fluid forms, and (II) nonlinear field-dependent conductivity (FDC) materials, where (I) is proposed as a substitute for silicone gel in high voltage, high power UWBG β-Ga2O3 devices. For electric field control within the module, new insulation systems will be innovated by applying a combination of nonlinear FDC materials as layers on high electric stress regions, mentioned above in (II), and using geometrical techniques to reduce electric fields to control partial discharges. It is a highly coupled and interconnected device-packaging project where new packaging methods are tailored to the ampere-class multi-kV β-Ga2O3 device targeted in this project. Further, the packaging will be extended to the device level, where, for example, substrate thinning will be examined in combination with packaging techniques at the module level. As a result, the project will develop an electro-thermal, device-package codesign framework that allows physical insight into the device-package interdependencies and speeds up the design of power modules that maximize the potential of emerging UWBG power devices. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-10

Reversible solid oxide cells are devices that can switch between two opposite operating modes for hydrogen production and power generation. These devices can potentially revolutionize the way hydrogen is made. Despite the promise, though, use of these devices faces significant challenges due to fast degradation of the cell under prolonged operation. This Faculty Early Career Development (CAREER) award supports research aiming to understand the complex degradation mechanisms within reversible solid oxide cells. By overcoming these challenges, the technology can enable cost-effective use of hydrogen as a clean fuel in industries and the heavy-duty transportation sector. Utilizing hydrogen as a long-term energy storage solution, it also promotes the integration of renewable energy sources into the grid. This research spans multiple disciplines such as solid mechanics, electrochemistry, and advanced imaging. In alignment with UMass Lowell's mission as a Minority Serving Institution, the project seeks to encourage inclusivity by engaging underrepresented groups in energy engineering research and education, fostering a more diverse and inclusive workforce in the field. The rapid degradation in reversible solid oxide cells during electrolysis mode, caused by delamination failure at the oxygen electrode/electrolyte interface, is commonly associated with the buildup of oxygen partial pressure. Significant scientific challenges persist in fully comprehending the mechanisms responsible for the reduced degradation observed under reversible modes, especially with a bilayer oxygen electrode configuration. This research aims to bridge the existing knowledge gap by investigating the intricate interactions between mechanical and chemical stresses at the oxygen electrode-electrolyte interface under dynamic operating conditions, utilizing integrated mechano-electro-chemical approaches. The research will attempt to unravel the complex dynamics of crack initiation and propagation within 3D heterogeneous microstructures of oxygen electrodes from advanced full-field X-ray imaging technique. Through rigorous coupling of model and experiment approaches, the contributions of material variations and structural geometry modifications to performance improvements will be delineated. These findings will make a marked impact on the design of new oxygen electrodes and development of protocols for safe operation of reversible solid oxide cells. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-09

Nonprofit organizations play multiple and sometimes conflicting roles in a democracy: they represent many views but can also provide services at the direction of government. Public trust in charities has decreased over recent years in many countries, but little research explains why. This project utilizes The Trans-Atlantic Partnership call on Democracy, Governance and Trust (DGT) to study cross-sector opinions on trust and accountability. The research team seeks to understand the variation in regulatory approaches, interpersonal trust, and popular sentiment toward public-serving institutions using the mutual perceptions of four audiences: operating charities, foundations, government agencies, and the public. Findings will provide useful guidance for regulators, charity workers, and donors on the broader role of regulation and democratic participation in society by documenting obstacles within or across national boundaries. The study has four phases that allow for multiple levels of comparison between audiences, geographies, and conditions such as exposure to other groups. We also allow for opinions to change, which stands in contrast to most of the cross-sectional survey work that dominates the field. In the first phase, each team will conduct an extensive document review to create historical-institutional profiles, including the scope of the nonprofit sector and relevant socio-cultural norms such as institutional trust toward the nonprofit, private, and public sectors. In the second phase, teams will convene both single-audience focus groups and mixed-audience Delphi groups to gather novel data. The third phase will produce practice-oriented reports, while the fourth phase involves academic book and article production and contains a conference event that serves not only to share knowledge, but serves as another round of data gathering to document opinion shift and further study the influence of peer learning on trust, accountability, and governance. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-09

Nontechnical Description The exceptional electronic and optical properties of metal halide perovskites (MHPs) have drawn great interest for their use in electronic devices such as solar cells and light-emitting diodes. However, the stability of devices based on perovskites is an outstanding problem. Low-dimensional, fully inorganic perovskites have the potential to address this issue. This project develops heterostructures made from light-emitting perovskite nanocrystals embedded in a bulk perovskite matrix. The matrix provides efficient charge transport, and the nanocrystals emit bright and saturated red, green, and blue light. Furthermore, these structures show enhanced stability in operating devices. The work involves understanding the synthesis of novel nanocrystals, processing semiconductor blends, and developing structures for highly efficient and stable devices. This outcomes of this project impact the development of energy efficient electronics, including displays, flexible electronics, and solid-state lighting. Students will gain training that spans physics, chemistry, photonics, and materials science and engineering. Furthermore, the project fosters connections to industrial projects and international collaborations, providing pathways for future STEM careers. Technical Description Perovskite materials combine the advantages of both organic and inorganic emitters, with narrow emission bandwidth, high color tunability (by halide choice), facile chemical synthesis, and excellent charge transfer properties. The primary reason for instability, however, lies in the predominantly “soft” ionic structure of traditional (3D) materials with corner-sharing octahedra surrounded by the organic cations residing in the voids, thus providing insufficient stability. To address these problems, the research team proposes to develop and utilize inherently more stable, zero-dimensional (0D) perovskite nanocrystals (NCs) without corner-sharing connectivity. The higher degree of localization leads to much-improved stability. Our synthesis of highly emissive cesium lead bromide and cesium lead iodide 0D NCs and their seamless incorporation into the 3D perovskite injection matrix opens the possibility of creating stable, efficient white-light light emitting structures. For that purpose, the investigators will design perovskite light-emitting chemical cells employing polymer electrolytes and salts to invoke differential ion motion. When blended with perovskite film, smaller additive ions move in place of the perovskite ones, preserving the underlying structure of perovskite and allowing for an efficient charge injection from a single-layer device. This leads to low-voltage devices with record operational stability and lifetime. Specifically, the team will develop various methods of uniform or layer-by-layer blending of 0D perovskite NCs into a 3D matrix and conduct extensive optical and electrical characterization to develop structures for use in efficient devices. These strategies will ultimately open numerous avenues to explore low-dimensional perovskite materials for high-performance emissive devices. Importantly, this research on the interface of several disciplines will be a catalyst for many students to choose highly advanced fields of nanotechnology and will benefit them by providing diverse backgrounds in physics and materials science. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-09

Non-technical Abstract Two-dimensional (2D) semiconductors have been at the focus of condensed matter physics, driving the exploration of fundamental principles, novel quantum matter, and advanced functional devices. The moiré superlattices of 2D semiconductors allow for the emergence of diverse correlated-electron phases within a single device due to advantageous electric gate tunability yet suffers from challenges in achieving spatial inhomogeneity and reproducibility. Remarkably, naturally occurring crystalline multilayer graphene, a family of 2D semiconductors with electric-field tunable band gaps, provide an ideal and highly reproducible platform for studying similarly diverse correlated-electron phases within a single device, eliminating the need for moiré engineering. This project aims to investigate the emergent phenomena, phases, properties, and principles in this unique platform. The anticipated outcomes of this project are expected to benefit society through the research results and outreach initiatives. Its success will also illuminate the transformative potential of ultra-thin low-dissipation quantum chips in advancing next-generation information technology and sustainable energy solutions. Key educational endeavors integrated into the research include: (i) mentoring STEM undergraduate students from MIT, UTD, and REU programs, (ii) mentoring STEM undergraduate students from MIT, UTD, and REU programs, (iii) enhancing existing public outreach and K-12 education by developing demonstration module and lessons on 2D semiconductors. Technical Abstract Built upon the high material quality, electric gate tunability, and presence of strongly interacting electrons within highly reproducible rhombohedral graphene multilayers, this project aims to establish an exceptional platform focused on this distinctive family of crystalline 2D semiconductors. The goal is to investigate spontaneous symmetry breaking, demonstrate novel correlated phases of matter, and provide new insights into fundamental principles governing the intricate interplay between geometry, symmetry, topology, and interaction. This project leverages strong and complementary expertise in both experimental and theoretical domains, fostering an established collaboration. Successful implementation of the project will actualize the potential of rhombohedral graphene semiconductors in discovering, realizing, and optimizing a variety of exotic correlated phases of matter that are rare otherwise. These advancements will not only contribute to new knowledge in correlated electrons and semiconductor physics but may also pave the way for transforming next-generation information technology and sustainable energy solutions, utilizing ultra-thin low-dissipation quantum chips. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-09

Recent advancements in natural language processing (NLP) have led to a rapid increase in available text data, sparking research developments in precision medicine, economics, recommendation systems, and social science. While existing deep learning methods can predict outcomes accurately, it remains unclear how to disentangle, quantify, and use complex relationships among observed textual variables. Causal inference presents a solution for extracting trustworthy causal relationships and establishing counterfactual realities. This research project aims to develop statistical theories, methods, and algorithms to learn causal structure and establish causal identifications for text data. The project will impact various sectors, including the medical, financial, and health communities, promoting interdisciplinary collaboration. To tackle the challenges imposed by text data, the research project aims to solve the following tasks: (1) Establish a new approach to low-dimensional representation learning for text variables, with a primary focus on causal identifications; (2) Develop the textual causal structural learning and causal direction learning to identify the complex causal relationships between different text variables of interest, (3) Build a comprehensive analysis framework for average and heterogeneous textual causal effects that are able to accommodate textual features, textual actions, and textual outcomes, and develop their estimations with multiple robustness. (4) Construct an individualized online policy optimization framework tailored for text variables. During the involvement of the project, efficient computational algorithms that are designed to handle the challenges posed by large-scale and heterogeneous text data will be developed and implemented. In addition, the project will conduct software development for target applications in precision medicine and personalized recommendations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-08

The main goal of this project is to develop and deliver remote experiments utilizing cloud-based resources aimed at educating a broad audience of students and practitioners in hardware security. In the post-COVID era, it is imperative to develop online education platforms for remote training of both students and the workforce in the field of Hardware Security. Recent advances in this field and FPGA-based cloud servers have enabled an opportunity to move related experiments to an online format that only requires a standard computer and internet connection by the students. Teaching “hardware” security in a socially distanced format poses significant challenges. Essential experiments for teaching key concepts in hardware security necessitate multiple evaluation boards and physical equipment such as voltage supplies, oscilloscopes, multimeters, and function generators. To adapt these experiments for an online platform, the project will explore innovative methods to execute or emulate them using the cloud ecosystem. This project addresses a critical gap by developing a fully online hardware security training module accessible to students and professionals worldwide. This project proposes various comprehensive experiments testing different notions in hardware security. The framework will be designed for both undergraduate and graduate students in the electrical engineering, computer engineering, and computer science departments, leveraging courses developed by the PIs in their respective institutions. The proposed infrastructure includes preparing detailed experiments for instructors with walkthrough documents and organizing student assignments for independent completion. This setup supports not only teaching but also facilitates independent research upon assignment completion. Supplemented with video instructions, these experiments will constitute a comprehensive training module, equipping participants with the necessary skills and knowledge to address complex challenges in this emerging domain, thereby instilling preparedness and confidence. This award is co-funded by the NSF Improving Undergraduate STEM Education (IUSE: EDU) Program. The NSF IUSE: EDU Program supports research and development projects to improve the effectiveness of STEM education for all students. This project is further supported by the Secure and Trustworthy Cyberspace (SaTC) program, which funds proposals that address cybersecurity and privacy, and in this case, cybersecurity education. The SaTC program aligns with the Federal Cybersecurity Research and Development Strategic Plan and the National Privacy Research Strategy to protect and preserve the growing social and economic benefits of cyber systems while ensuring security and privacy. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-08

This RAPID award aims to continue to process, curate, and distribute plasma data from the operational DMSP spacecraft for the space science community during this critical period of 2024-2025 as we reach solar maximum. We are currently nearing the maximum of the 11-year solar cycle (with the peak expected sometime in 2025) as shown by the recent increase in the number and intensity of geomagnetic storms and activity. An example of this is the “Mother’s Day Storm” in May 2024 which was the largest geomagnetic storm to hit the Earth since 2003 resulting in auroras that were visible in most of the continental United States. These storms influence much of our civilian and military technological and infrastructure such as satellite communications, GPS, and power grid operations. The publicly available thermal plasma observations (densities, temperatures, flows, and composition from the SSIES-3 package) from the operational DMSP spacecraft were of only level 1 quality and contained large amounts of poor-quality data. Curating this database of the SSIES-3 ionospheric observations and providing it to the space science community will allow the full community to use these data in their own modeling and research efforts. These ionospheric data are necessary as inputs for numerous researchers and groups using them to develop a better understanding of the behavior and dynamics of the ionosphere. The curated data will be delivered to the U.S. NSF-funded Madrigal and NASA SPDF data centers for public distribution and use. Prior to the team’s previous work, the publicly available thermal plasma observations (ion densities, temperatures, flows, and composition from the SSIES-3 package) from the operational DMSP spacecraft (F16, F17, and F18) were of only level-1 quality and contained large amounts of poor-quality data that were not flagged as such for the end user. During the past four years, the team updated the data reduction code and reprocessed over 25 satellite-years of SSIES-3 data from 2003 onward producing a level-2 quality dataset along with quality flags on most of the parameters. These improved level-2 data were delivered to the U.S. NSF-funded Madrigal and NASA SPDF data centers for public distribution and use. This RAPID award will fill the gap in the availability of these data products during the solar maximum just when they are most critically needed by the space science research community. The team will continue to produce these data for all three spacecraft for all of 2024 and most of the 2025 period. These level-2 data will be delivered to the data centers for the space science community’s use. As time permits, the team will work backwards from 2022 to fill in some (but not all) of the gap in the data that still needs to be reprocessed. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-08

This project is funded from the Research Experiences for Undergraduates (REU) Sites program in the Directorate for Social, Behavioral and Economic Sciences (SBE). It has both scientific and societal benefits in addition to integrating research and education. Currently, the majority of developmental research is conducted by and focused on monolingual individuals from middle-to-high income households. To address the critical need to expand developmental theories to incorporate underserved and underrepresented populations, this project will train students from these understudied populations to become researchers who can provide rich and needed perspectives on developmental science. Through participation in the University of Texas at Dallas’s year-long REU, 30 students (10 per year) will learn the necessary tools to conduct high-quality research with real-world applications with families that are challenging-to-study, vulnerable, and underrepresented in the developmental science literature. Thus, the project will advance the field of developmental science by increasing diversity in both the researchers and the children who are researched. Students will be drawn primarily from the Dallas/Ft Worth area, which compared to other similarly large cities in the U.S., has few research intensive 4-year institutions of higher learning, and few opportunities for students to gain high quality, first-hand research experiences. At least 50% of the students will be enrolled at one of Dallas College system’s 7 community campuses, and the remaining students will be enrolled at The University of Texas at Dallas. A large number of diverse students attend the local community colleges and UTD. In the fall, students will acquire skills needed for culturally-responsive research by taking an active role in the Play With Me program. Play With Me, a unique and important aspect of this REU’s research experience, is a free 12-week parent/child (ages 0-3 years) community-based outreach program created by UTD developmental scientists and specialists based on playful-learning research and directed by developmental specialists of UTD’s Center for Children and Families. With mentorship from UTD faculty, students will formulate new theoretically-driven research questions within ongoing data sets influenced by their Play With Me experience. In the spring, a faculty mentor team will guide students as they collect and analyze data from families recruited from Play With Me. In the summer, students will learn to disseminate findings to academic (journals, conferences) and non-academic (newsletters, community outreach) audiences. Fellows will also participate in brownbag discussions on professional development, including how to identify and apply to graduate schools, prepare for the GRE, and identify academic career options. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-07

Young children’s language development is crucial for their later academic success. During the preschool years, teachers contribute to children’s language development by engaging in conversations with them. However, it is not clear how often preschool teachers have language interactions with children, or if every child in the class receives opportunities to talk with teachers. This project uses innovative sensing technology tools to examine teachers’ language interactions with preschoolers over the course of a school year. The project employs advanced speech processing algorithms that automatically analyze language data, providing a more detailed understanding of how language is used in the classroom than has previously been possible. The results provide insights into how preschool language experiences shape language development and whether children’s language experiences differ based on their background characteristics, such as their standardized English language skills, temperament, or gender. The eventual goal of this work is to provide teachers with concrete, actionable data on how to enrich language experiences in preschool and lay a strong foundation for future reading success. The study seeks to answer the following research questions: How often do teacher language interactions occur in preschool classrooms? Which children in a classroom are involved in fewer or briefer teacher language interactions, as compared to their peers? Do teacher language interactions change over the course of the year? Are teacher language interactions associated with growth in children’s English language skills? Monthly data are collected in preschool classrooms using sensing technology tools to address these questions. These sensing technology tools allow for the identification of language interactions, or moments when a teacher and child are in proximity to each other and are talking. Information is also collected on children’s language skills, temperament, teacher-child relationships, and demographic information. The outcomes of this project are expected to inform models of language development in preschool-age children and provide fine-grained data that can be used by educators to equitably support language development in preschool settings. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-07

VR (Virtual Reality)-based immersive and interactive environments can be a great resource for learning and training, especially for concepts that involve safety aspects by interacting with inflammable or breakable objects. The tedious nature of developing these VR experiences continues to be a limiting factor for VR becoming mainstream. The main goal of this project is the design and development towards an eventual innovative infrastructure, SMILE (Scan to Multi-sensorial Interactive Learning Environment), focusing on: (1) nearly automated construction of VR environments that mimic real-world indoor scenes; (2) interactions with virtual objects involving multiple senses such as touch, visual, aural, and smell. The SMILE infrastructure will undergo rigorous software testing to ensure safe interactions before being deployed to support Internet-scale collaboration among users. In the exploratory phase of this project, the team will develop a set of tools and software modules for the following purposes: (a) Realistic construction of VR environments to represent real-world space(s) as closely as possible; (b) Natural interactions with the constructed VR environments using multisensorial devices; (c) Multimodal streaming to synchronize transmission and reception of multisensorial data over the Internet; (d) Testing and validating the SMILE software for robustness and safety. Virtual chemistry laboratory experiments for assessing the performance of SMILE will be designed and developed through research collaboration with the Dallas College, an HSI (Hispanic-serving institution). Through this collaboration, the SMILE project will have a significant impact on the under-represented student community. SMILE could be used for different learning environments, apart from the Chemistry lab case study, promoting education involving K-16. The project will involve women Ph.D. students, undergraduate, and under-represented community students. Open-source resources produced through the SMILE project include multi-sensory database of material properties for sensorial displays, Unity-based VR authoring tools, automated software testing tools for VR-based applications. These will be made available for a period of five years through the project website at: https://labs.utdallas.edu/multimedialab/. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-07

This project focuses on enhancing the safety and reliability of autonomous driving systems, making them more trustworthy in a variety of environments, including potentially hostile ones. The research will develop advanced detection and countermeasure techniques at the application, system, and hardware layers to help ensure that self-driving cars can make safe driving decisions even when faced with deliberate attempts to disrupt their operation. This involves creating robust deep-learning modules for driving decisions, enhancing the safety and security of automotive processors, and developing tools to detect and fix hardware issues in mission mode. By conducting extensive real-world testing with popular automotive benchmarks, the project aims to validate these innovations, ensuring they can be confidently adopted in everyday use. This comprehensive approach addresses current and future challenges in autonomous driving technology, paving the way for safer and more efficient transportation systems that the public can trust. The broader impact of this research extends beyond automotive technology to other critical systems such as space exploration, smart medical devices, and various Internet of Things (IoT) applications. With the deep learning market expected to grow to $24.5 billion by 2025, advancements from this project will play a crucial role in ensuring the safety and security of numerous technologies that impact daily life. The project’s findings will be widely disseminated through publications, software releases, and educational courses, contributing to the development of a skilled workforce in this rapidly evolving field. Further, the researchers will engage in educational outreach, including workshops and summer camps for students from diverse backgrounds, promoting inclusivity and inspiring future scientists and engineers. The researchers’ ongoing collaboration with leading semiconductor companies will also support translation into practice. The project not only advances technology, but contributes to societal well-being by making autonomous systems safer and more reliable, ensuring that the benefits of these innovations are widely accessible. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-06

Prior research and applications of voice conversion models have raised challenging problems that are both theoretical and use-inspired. Notable challenges include processing emotional speech and speech in noisy environments and generating speech that represents the characteristics and expressiveness of specific speakers such as personality traits, mood, prosody, and emotional state. These challenges are exacerbated by a limited availability of data. Improving such capabilities will have a wide range of social impacts ranging from giving natural voice to patients who have lost it to rendering comprehensible and speaker faithful renderings of old poor quality recordings that have become hard to understand to generating seamless speech translations in real time communications while staying faithful to the voice characteristics of the speaker. To address these challenges, the project proposes to explore and expand theories about speaker identity, emotion, and expressiveness in challenging conditions. Practically, this means studying how factors like background noise, emotions such as stress, cultural differences or other idiosyncratic ways of speaking affect a system’s ability to recognize and render faithfully the speech of a specific individual. This work will enable a second aim of this project which is to create voice technology that can be used for safeguarding ethical and responsible use of voice generation. Sophisticated voice conversion techniques can be used to detect and prevent spoofing and other fraudulent activities and make it challenging for unauthorized users to mimic or imitate target speakers. Besides security and defense other areas that will benefit from this project include security and defense, accessibility and healthcare assistive technologies, medical voice preservation, speech therapy and rehabilitations as well as entertainment and gaming. This award aims to develop novel algorithms utilizing deep learning techniques to advance voice conversion models with the ability to represent faithfully the characteristics and emotional states of individual speech. The project includes the following key areas of research. The first research target is to explore learning speaker identity and emotion representations for robust voice conversion with self-supervision. By investigating joint representations, this project seeks to develop a deeper understanding of how speaker characteristics and emotions can be effectively transformed. The second research target is to investigate voice conversion solutions for challenging conditions such as noisy environment, emotional speakers, and limited training to enhance the expressiveness and naturalness of the converted speech. The third research target is to investigate novel deep learning techniques for the detection of synthetic voices and joint training strategies to further improve voice conversion performance and evaluation. By exploring the synergies between transformation and detection of synthetic voices, this project has the potential to significantly impact society with a) accurate and expressive voice-based applications and b) applying the same techniques to detect when speech is naturally occurring or synthetic for the prevention of spoofing. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.