Although significant advancements have been made in Unsupervised Domain Adaptation (UDA), existing methods are typically validated on curated public datasets that do not adequately represent the complexities of real-world applications, such as imbalanced (long-tailed) label distributions and simultaneous domain and label shifts. To bridge this gap, we introduce the Target Distribution Agnostic Adaptation Network (TDAAN), a novel framework designed to facilitate robust adaptation from source to target domains under significant label shifts. Additionally, we present a new domain adaptation dataset, MS-DA, which focuses on marine species and incorporates natural domain and label shifts, a feature sorely lacking in current domain adaptation research. Our experiments show that TDAAN not only significantly outperforms the baseline UDA method but also surpasses the performance of leading UDA methods on the MS-DA dataset. Remarkably, TDAAN maintains competitive performance on standard UDA benchmarks, proving its efficacy even in scenarios with minimal label shifts. These results position TDAAN as a superior method for UDA, particularly in real-world applications characterized by complex and diverse data distributions. Our code is available at https://github.com/SEFSC/FATES-ATI-DomainAdaptationLabelShift.

In recent years, remote laboratories have become integral to modern education, offering flexibility and accessibility compared to traditional, in-person labs. Integrating AI-powered assistance into remote labs has the potential to give them an edge by providing personalized learning experiences. This paper explores an innovative approach to promoting independent learning and critical thinking by embedding AI-driven support, using OpenAI's GPT-4 model, into a remote Field Programmable Gate Array (FPGA) laboratory. Through a web-based code editor, students write SystemVerilog programs and receive tailored assistance from the AI, while their designs are deployed on a Terasic DEl-SoC FPGA development board with real-time feedback via a live camera feed. The study, which involved students from an advanced digital design course interacting with the AI assistant, revealed strong engagement and positive feedback. Preliminary results indicate that AI-powered guidance can meaningfully boost student involvement, providing a scalable and effective framework for fostering active learning in engineering education.

In this Research-to-Practice paper, we detail the development and implementation of a 3D Digital Twinning simulator augmented with Virtual Reality (VR) capabilities, specifically designed for an embedded systems curriculum. With the rise in technology-enhanced learning tools, there has been a marked shift towards creating robust platforms that allow students to interact more deeply with their coursework. Digital Twinning and 3D virtual simulations have emerged as significant contributors to this shift. The challenge in most design-oriented courses lies in the integration of multiple tools and hardware components. Students often find themselves investing significant time in the setup and configuration of these components, time that could otherwise be devoted to the core learning objectives. Our 3D simulator was developed to alleviate this burden. It intends to offer a clear, intuitive visualization of tasks, allowing students to focus on designing and testing their implementations. This simulator does not just represent a theoretical concept but provides a realistic, interactive environment replicating real-world scenarios. One of the central case studies highlighted in this paper is the transition from a traditional FPGA-based digital design assignment-a car parking lot scenario with switches and breadboards-to an equivalent, but more immersive, 3D virtual simulation. This redesigned model boasts seamless interfacing capabilities with a remote FPGA lab, further bridging the gap between theoretical learning and practical application. To gauge the effectiveness and utility of our simulator, we conducted an anonymous survey among the participating students. The survey aimed to capture feedback on the usability, intuitiveness, and overall educational value of the Digital Twinning simulation. Preliminary findings from our study suggest a positive trend: Digital Twinning, combined with the immersive properties of VR, has the potential to significantly improve engagement, comprehension, and performance. This research underlines the evolving dynamics of education, where traditional methods are gradually being complemented, if not replaced, by technology-enhanced approaches. The results presented in this paper strongly suggest the potential benefits and adaptability of Digital Twinning in the contemporary educational landscape.

Unsupervised Domain Adaptation (UDA) aims to transfer knowledge learned from a labeled source domain to an unlabeled target domain, mostly through learning a domain invariant feature representation. Currently, the best performing UDA methods use category level domain alignment to capture fine-grained information, resulting in significantly improved performance over global alignment. While successful, category level UDA methods suffer from the unreliable pseudo-labels for target data. In this paper, we propose an UDA approach with teacher-student learning where the teacher network is used to provide more reliable target pseudo-labels for the student during training. Furthermore, we use a progressive mixup augmentation strategy which generates intermediate samples that become increasingly target-dominant as training progresses. Aligning the source and intermediate domains allows the model to gradually transfer fine-grained domain knowledge from the source to the target domain while minimizing the negative impact of noisy target pseudo-labels. This progressive mixup augmented teacher-student (PMATS) training strategy achieves state-of-the-art performance on two public UDA benchmark datasets: Office-31 and Office-Home.

This Research Full Paper builds on a prior study that compared overall student performance between in-hand versus remotely accessible hardware in digital design courses. The COVID-19 pandemic necessitated a global educational shift to emergency online learning that led to rethinking the delivery of engineering labs. The prior study showed that, amidst pandemic-necessitated online learning, student understanding was not impeded by the incorporation of remotely accessible hardware into the course curriculum; rather, using remote hardware resulted in similar or better learning outcomes. In this paper, we analyze the remotely accessible hardware lab through the lens of equity, investigating the student perspective on equitable access and the remote lab experience. The study accomplishes this goal by surveying students of a junior-level digital design course who use a remotely accessible hardware lab for completing their assignments. The survey aims to determine the factors deemed important by today’s learners – those who have experienced remote learning for approximately two years of their educational careers – when considering equitable access and remote labs. Survey questions utilized the multiple-choice, semantic differential scale, and Likert scale formats for quantitative analysis as well as inductive coding of freeform responses for qualitative analysis. Initial findings from the survey are the key considerations of the surveyed students which include Factors of the Remote Experience (FREs) and Factors of Equitable Access (FEAs). FREs and FEAs specifically relate to the Student’s Access to Electronic Devices, the Student’s Environment Outside of Class, the Student’s Schedule, the Student’s Internet Quality, the lab’s Learnability, the lab’s Web Interface Design, the lab’s Convenience, the lab’s Overall Positive Experience, the lab’s Ease of Use, the lab’s Internet Quality, and the lab’s Affordability. Rooted in the online learner’s experience, these results contribute to an improved understanding of how students perceive equitable access to engineering education which shall guide better-informed advancements in the field in a post-pandemic world.

This Research to Practice Work-In-Progress paper presents a virtualized breadboard solution for FPGAs and ARM microcontrollers in remote laboratories. The circumstances that rose amidst the COVID-19 pandemic demonstrated the vulnerability of current engineering education practices, particularly in dealing with hardware resources. Pivoting to the emergency online instruction presented challenges to the traditional practices in delivering hands-on engineering labs, which necessitated a solution that handles hardware prototyping without compromising creativity and instruction. One vital aspect of the embedded systems learning experience is ensuring students and faculty members alike have opportunities to learn and build custom prototyping circuits that interact with microprocessors on breadboards. In this paper, we build on the prior work that our group implemented on using virtualization to interface a virtual breadboard with physical hardware through web applications. Our previous work was limited to interfacing with one particular kind of hardware, designed to explore the capabilities of fundamental transducers and actuators that interface with hardware I/O pins. In hardware engineering practice, however, designers are not constrained by a single microprocessor selection to control their system and designs and are not limited by the type of transducers and actuators that provide the external circuit functionality. This paper presents a solution by scaling the existing virtual breadboard research to support FPGAs and ARM microcontrollers and intermediate logic gate integrated circuits for practical use in engineering curriculums. Providing this increased selection of supporting hardware helps facilitate student learning and simulates hardware development in an industrial setting. Due to the rising popularity of FPGAs and ARM microcontrollers in industry and in education, we expect that our solution will serve a larger audience through this broader selection of supported hardware. Our solution virtualizes the breadboard prototyping experience without sacrificing the nature of real-time embedded systems by taking the user prototyped inputs and outputs and directly programming the functionality of the surrounding system to physical hardware. This balance between a virtualized interface and physical hardware implementation preserves a hardware curriculum embedded systems engineering education and brings a promising solution to expand the scalability and accessibility of engineering labs.

Currently there are little objective parameters that can quantify the success of one form of prostate surgical removal over another. Parameters such as the percent of coverage and depth of extra-capsular soft tissue removed with the prostate by the various surgical approaches can aid in quantifying prostate surgery success. Previously, we presented visualization methods as well as measurement algorithms to determine extra- capsular tissue coverage. In this paper, we present a modification to the coverage algorithm where we changed the way of mapping the reconstructed model as well as applying different neighborhood check approaches.

Currently there are little objective parameters that can quantify the success of one form of prostate surgical removal over another. Accordingly, at Old Dominion University (ODU) we have been developing a process resulting in the use of software algorithms to assess the coverage and depth of extra-capsular soft tissue removed with the prostate by the various surgical approaches. Parameters such as the percent of capsule that is bare of soft tissue and where present the depth and extent of coverage have been assessed. First, visualization methods and tools are developed for images of prostate slices that are provided to ODU by the Pathology Department at Eastern Virginia Medical School (EVMS). The visualization tools interpolate and present 3D models of the prostates. Measurement algorithms are then applied to determine statistics about extra-capsular tissue coverage. This paper addresses the modeling, visualization, and analysis of prostate gland tissue to aid in quantifying prostate surgery success. Particular attention is directed towards the accuracy of these measurements and is addressed in the analysis discussions.